<%BANNER%>

Scientific Applications of the Mobile Terrestrial Laser Scanner (M-TLS) System

Permanent Link: http://ufdc.ufl.edu/UFE0021101/00001

Material Information

Title: Scientific Applications of the Mobile Terrestrial Laser Scanner (M-TLS) System
Physical Description: 1 online resource (133 p.)
Language: english
Creator: Fernandez-Diaz, Juan C
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: geosensing, laser, lidar, mapping, mtls, remote, scanning, sensing, terrestrial
Civil and Coastal Engineering -- Dissertations, Academic -- UF
Genre: Civil Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Terrestrial Laser Scanners (TLS) are mapping instruments composed of a Laser rangefinder and an optical-mechanical system used to steer the laser beam across the surface of interest. The University of Florida Geosensing Engineering and Mapping (GEM) Research Center is working towards developing a Mobile Terrestrial Laser Scanner (M-TLS) system. The core of the M-TLS is a commercial 2-axis ground based laser scanner which is integrated to a mobile telescoping, rotating, and tilting platform which provides up to 6 degrees of freedom for performing scanning operations. A scanner built-in 6 megapixel digital camera and a digital video camera provide the M-TLS moving and still imaging capability. At the final stage of development the M-TLS positioning and navigation system will include a differential GPS array, tilt sensors and an inertial measuring unit which will allow data collection and georeferencing in both static and dynamic modes. The M-TLS laser scanner is capable of generating 3D spatial and multispectral data sets. A typical dataset is composed of a cloud of millions of points for which 3D coordinates, laser intensity and/or RGB information are available for each one. Data can be collected in a range from 3 m to 1500 m for a target with an 80% reflectivity or 3 m to 350 m to targets with a 4% reflectivity. The laser operates at a wavelength of 1535 nm, with a pulse width less than 10 ns and energy of less than 10 microjoules. The sampling separation can be adjusted down to 0.00115degree, and the scanning speed is 2,000 points per second. The M-TLS is a unique tool that enables GEM researchers to acquire high density point clouds from an advantageous terrestrial geometry, being a very valuable complement for Airborne Laser Scanner data sets. The applications of the M-TLS data sets are numerous in both the fields of science and engineering. Tested applications by the GEM center include urban mapping, as-built surveying, building damage assessment, bridge load analysis, forestry metrics extraction, beach erosion mapping, paleontology and archeology dig mapping, structural geology mapping, forest fire fuel estimation, soil spatial characterization and vehicle 3D modeling. This thesis centers around the novel applications of the M-TLS to specific scientific problems: thoroughly analyzing the applications to beach erosion hot spot mapping, soil roughness metrics extraction, and forestry metrics extraction.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Juan C Fernandez-Diaz.
Thesis: Thesis (M.S.)--University of Florida, 2007.
Local: Adviser: Shrestha, Ramesh L.
Local: Co-adviser: Slatton, Kenneth C.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021101:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021101/00001

Material Information

Title: Scientific Applications of the Mobile Terrestrial Laser Scanner (M-TLS) System
Physical Description: 1 online resource (133 p.)
Language: english
Creator: Fernandez-Diaz, Juan C
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: geosensing, laser, lidar, mapping, mtls, remote, scanning, sensing, terrestrial
Civil and Coastal Engineering -- Dissertations, Academic -- UF
Genre: Civil Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Terrestrial Laser Scanners (TLS) are mapping instruments composed of a Laser rangefinder and an optical-mechanical system used to steer the laser beam across the surface of interest. The University of Florida Geosensing Engineering and Mapping (GEM) Research Center is working towards developing a Mobile Terrestrial Laser Scanner (M-TLS) system. The core of the M-TLS is a commercial 2-axis ground based laser scanner which is integrated to a mobile telescoping, rotating, and tilting platform which provides up to 6 degrees of freedom for performing scanning operations. A scanner built-in 6 megapixel digital camera and a digital video camera provide the M-TLS moving and still imaging capability. At the final stage of development the M-TLS positioning and navigation system will include a differential GPS array, tilt sensors and an inertial measuring unit which will allow data collection and georeferencing in both static and dynamic modes. The M-TLS laser scanner is capable of generating 3D spatial and multispectral data sets. A typical dataset is composed of a cloud of millions of points for which 3D coordinates, laser intensity and/or RGB information are available for each one. Data can be collected in a range from 3 m to 1500 m for a target with an 80% reflectivity or 3 m to 350 m to targets with a 4% reflectivity. The laser operates at a wavelength of 1535 nm, with a pulse width less than 10 ns and energy of less than 10 microjoules. The sampling separation can be adjusted down to 0.00115degree, and the scanning speed is 2,000 points per second. The M-TLS is a unique tool that enables GEM researchers to acquire high density point clouds from an advantageous terrestrial geometry, being a very valuable complement for Airborne Laser Scanner data sets. The applications of the M-TLS data sets are numerous in both the fields of science and engineering. Tested applications by the GEM center include urban mapping, as-built surveying, building damage assessment, bridge load analysis, forestry metrics extraction, beach erosion mapping, paleontology and archeology dig mapping, structural geology mapping, forest fire fuel estimation, soil spatial characterization and vehicle 3D modeling. This thesis centers around the novel applications of the M-TLS to specific scientific problems: thoroughly analyzing the applications to beach erosion hot spot mapping, soil roughness metrics extraction, and forestry metrics extraction.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Juan C Fernandez-Diaz.
Thesis: Thesis (M.S.)--University of Florida, 2007.
Local: Adviser: Shrestha, Ramesh L.
Local: Co-adviser: Slatton, Kenneth C.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021101:00001


This item has the following downloads:


Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101129_AAAAAI INGEST_TIME 2010-11-29T21:26:45Z PACKAGE UFE0021101_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 1053954 DFID F20101129_AAAFHI ORIGIN DEPOSITOR PATH fernandezdiaz_j_Page_040.tif GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
e1cfabc5d32172e5ee705aee8cbb0591
SHA-1
da8d6477368bcfea77a74dc8d36f566741b8a782
41872 F20101129_AAAFGT fernandezdiaz_j_Page_072.pro
7824ac3d9fd6bd64b48b7ea445bf6165
44a1021e7d3b547d427c0a393256d033558c2d7c
105502 F20101129_AAAFHJ fernandezdiaz_j_Page_006.pro
2ab728ca8981a99103c097e56b4fc962
1dfd2e657506a75496b78cfad4302b24c98f1e33
52007 F20101129_AAAFGU fernandezdiaz_j_Page_050.pro
0eb3c57a1a88113e15abff15666d1e54
7d71c1b7464bc42043e4e9fe810788ea7f335e03
6648 F20101129_AAAFHK fernandezdiaz_j_Page_039thm.jpg
c35ed34780c075e778a7ac5d5caa9493
e811475b3f8ee55ebd24153c5d57da458c9976e0
75007 F20101129_AAAFGV fernandezdiaz_j_Page_067.jpg
359833baa6d5f2bc5e53359c3a0773e2
68e798c7110ed4042ec4d06cff51c4e6eae1d2c0
25846 F20101129_AAAFHL fernandezdiaz_j_Page_017.QC.jpg
30bc63e403c9fde81a54924c3b643296
ec7f615c4ed657b7a868649368177b3efd75dc14
54895 F20101129_AAAFGW fernandezdiaz_j_Page_132.pro
d2ac090b6a270fa1654d03d6c0b79346
8eb67fd0d8157c0f4a0acaf374dbeccf69f1f776
F20101129_AAAFHM fernandezdiaz_j_Page_001.tif
e3078f9ff852157257193330275d53e1
8df959a47fed2a6093e1ba3bc01f31c42e06d1a3
1051973 F20101129_AAAFGX fernandezdiaz_j_Page_114.jp2
96c63162d6a52cb67da915df6ecd2fb4
631021eeabc398e53c240987e616b6bf3b490efa
1051930 F20101129_AAAFIA fernandezdiaz_j_Page_068.jp2
f02d3b128f773c9c6d2aafefcddca25a
2f68bf2d7c783d16c4983eadd2dbd13c318c7748
F20101129_AAAFGY fernandezdiaz_j_Page_053.tif
389edb4b0f5e67a1df4934fa6d4551a4
d00406a33546424e925721d9ddee535cdd814961
1051985 F20101129_AAAFIB fernandezdiaz_j_Page_090.jp2
20bb220bb7dc3a567c110a90d2e5a9b2
1b1c3f96743d22cc379a4b94e3a7733933fd3931
110055 F20101129_AAAFHN fernandezdiaz_j_Page_116.jp2
2fef66eb135ea400c31d404f505c3e04
f714fabf6a5b16f4de46b112eaeb2fe98bb5a429
25271604 F20101129_AAAFGZ fernandezdiaz_j_Page_042.tif
7aa8b5ecec60bc9f8ae871ebdaaffbd8
697bbfcf9f994d66ba2854e7135cc03ab55b2ab4
29623 F20101129_AAAFIC fernandezdiaz_j_Page_015.pro
f3fa0b6e46b96f93e2f55f939bc5796e
482c427bb4a88069f2ac6f2416c50bcec0bdfdcd
25011 F20101129_AAAFHO fernandezdiaz_j_Page_001.jp2
7bdffd94aa2df095bf4475bc2e541b56
6956180c27907dbf9a9c1d54a55bae66f6310049
19566 F20101129_AAAFID fernandezdiaz_j_Page_113.QC.jpg
6ed68c097005c48a31e8a9d4dcebce3b
1669517f16a35c70010f18cb3ffab08190b02462
64999 F20101129_AAAFHP fernandezdiaz_j_Page_125.pro
24f03be3f4bcceabb18f3961854e225d
e54cabbcff557fa5d1726a02ae0c40406ff75bff
57642 F20101129_AAAFIE fernandezdiaz_j_Page_043.pro
950cb7063a99b956d8f927739fbd1f17
edc631e25dc7f923ea3294bf73959b8138423195
77266 F20101129_AAAFHQ fernandezdiaz_j_Page_054.jpg
8cec98c602446c0bf95c2fe1f3434af1
98414242b36e8a5e974c9650e00ffff41233a038
F20101129_AAAFIF fernandezdiaz_j_Page_012.tif
723bd94842a3fd58541376f3a04bfb78
cb7619cdd59f2890d7be9c5398046781ebc18e82
1912 F20101129_AAAFHR fernandezdiaz_j_Page_058.txt
40392eee3d41e7d8d54d49d269a0069e
7d67e4ee38ce5e2d2e9867a1200ee127d1391b90
F20101129_AAAFIG fernandezdiaz_j_Page_025.tif
eae38a170a9e7584fd97a7a3d821045e
090f115b675336fdf51bf884a662e11c077273f3
23914 F20101129_AAAFHS fernandezdiaz_j_Page_116.QC.jpg
d50a0cd6dc93817757375745e1b1e879
528eefd13fa286051920e1ab7bf780310f745e62
24462 F20101129_AAAFIH fernandezdiaz_j_Page_039.QC.jpg
57467ffb0aa14e35077a29a28f6904a2
8df14010ed6ba2c5a8d47aa3e4764d7fab8d02a1
1843 F20101129_AAAFHT fernandezdiaz_j_Page_033.txt
2b521a0fd1cca455b98b80168914f59c
70fe417d7649e71813412825d24323a455733cc9
86369 F20101129_AAAFII fernandezdiaz_j_Page_077.jpg
df42ec9b8def47a3cab77378c11ef6c1
03871f758732ad3753f59ced53b7da27d774dd16
F20101129_AAAFHU fernandezdiaz_j_Page_036.tif
fd2faa9de9f711869dda99e21d59f96d
b138a7f69c30c7a751432f18fcf0bc620f2c802b
1051977 F20101129_AAAFIJ fernandezdiaz_j_Page_129.jp2
d2040780e3da61992f1a3a31ce702188
5319655e609e5239b1753c7c57c65812b2924244
107834 F20101129_AAAFHV fernandezdiaz_j_Page_061.jp2
f4e0f4017b634d4167bb3c22e52ab8ff
10af898fbed3b1b38f450b183e2a8b3a227724f5
73533 F20101129_AAAFIK fernandezdiaz_j_Page_047.jpg
41e26e22359f9a4e86ffb0155c14cc75
ed7260346aa9760a0737561038086244d3f8fc23
1051963 F20101129_AAAFHW fernandezdiaz_j_Page_081.jp2
525a8018c1853ab3de453b7ac68ac888
14edda5b9e2037ff7cd40faf6c6a95085310f4a4
1051935 F20101129_AAAFIL fernandezdiaz_j_Page_092.jp2
68782ef80fd804d16f261aa25778c0bb
e04279be06e8aafe9f64154d15231f4812ea3bc6
74121 F20101129_AAAFHX fernandezdiaz_j_Page_012.pro
df848d7a9d315ab16eee288a8dede88e
e80a84e4ba9bbca3d6cb303c4187d9a2aec38d81
5721 F20101129_AAAFJA fernandezdiaz_j_Page_083thm.jpg
e74477cd48740a90b597d7b2cdfef5b7
00dd0c674c8b530ffc14c38cbdd5e7ab3067a254
6914 F20101129_AAAFIM fernandezdiaz_j_Page_040thm.jpg
9e21b6244605739c95d8446af6da7ea1
257d1abca3afa13ca7acbd5c3a6ee93e67359079
31276 F20101129_AAAFHY fernandezdiaz_j_Page_033.pro
31eab8949459f55f58b79a7e43914935
c3fdbe28b44aabb1c3e1f7862fadf4ea62763800
F20101129_AAAFJB fernandezdiaz_j_Page_116.tif
9365459117017c338c322ef0e0352d06
a402e1b4fd03b9b25ecf7c82c384fb24808c3566
F20101129_AAAFIN fernandezdiaz_j_Page_037.tif
ebe79678df549da64763b87bd6e5f899
627f50af8004311e13d2465bf34788fba6d874fe
6860 F20101129_AAAFHZ fernandezdiaz_j_Page_101thm.jpg
f1cc031f61680643b43779aea36c0bb4
fe293b6f2cb726ab8d9608fd38266471b538dd3b
2116 F20101129_AAAFJC fernandezdiaz_j_Page_062.txt
936375b3e797990f2bdac9edea40af89
4166eeb59fbd16edcaff48b81366f9ee3ada8b78
24666 F20101129_AAAFJD fernandezdiaz_j_Page_104.QC.jpg
eed0917a1002addd59b7e92b781e0591
e668c507c28c1b110083d222f54471e020f974f0
24733 F20101129_AAAFIO fernandezdiaz_j_Page_062.QC.jpg
25b8cbb93d95f98923d706ae9f395f64
1e7829b7e7caf71a6eb047f7914616e99782c8fb
6990 F20101129_AAAFJE fernandezdiaz_j_Page_071thm.jpg
9bec307b42c763ddefd43121a0defd55
910920aeff68a71127034e3aeb0027df1e9a3c01
66083 F20101129_AAAFIP fernandezdiaz_j_Page_125.jpg
37daab73ad2c2316ba6388fb7e3e9316
a8474ba1537a0bab586d8c626dea40b1aa38b6e0
95081 F20101129_AAAFJF fernandezdiaz_j_Page_059.jpg
3894f8d40e70ee106cf0dd3dfa92203a
ec2fd1b500e2963c12874065f47369829cf77596
814070 F20101129_AAAFIQ fernandezdiaz_j_Page_121.jp2
93dab66b6762db2a7d0881b09098d668
50ddd35940b6cd8d49f056bf3b7005b1cd36864c
24153 F20101129_AAAFJG fernandezdiaz_j_Page_029.QC.jpg
b21f4da394da39add91dcbac1bb194ef
0b2d508127e56090a89628189b451695b2038448
26411 F20101129_AAAFIR fernandezdiaz_j_Page_051.QC.jpg
5126182ae5ebc14b4ed41f9ddf468d47
43d0786458a8f3748a7c57dd1712f64184c694ae
6150 F20101129_AAAFJH fernandezdiaz_j_Page_124thm.jpg
77c255be8278c5b905fb11b0dabdbc5f
5f453b48e50a88dc4dd53a9ee7371ee5c65a5b65
79038 F20101129_AAAFIS fernandezdiaz_j_Page_040.jpg
f50dae5e80b990b5e576c4c73c6d0d7c
dcd30f0f2c4721d3f05b5f2cd8b3885e2051021d
27309 F20101129_AAAFJI fernandezdiaz_j_Page_072.QC.jpg
a7656f666fa8e0ea9ce86a644f67c632
759f01b15f36877952a64f694f18d5ed616c8b2d
1051969 F20101129_AAAFIT fernandezdiaz_j_Page_072.jp2
0729365118087b4d5458c9cd2bd1d0dd
bd44944dd5938eeeaef4ebeda3efd5d4ec5aa369
F20101129_AAAFJJ fernandezdiaz_j_Page_106.tif
2a941c7757ce8109bdcfb758e4f360f7
9fdab6b0826d0066d2888a68f0a665031dee01dc
7636 F20101129_AAAFIU fernandezdiaz_j_Page_046thm.jpg
f9414d857880d3ca188a7fae9fcb70b9
75854652c92dc53a36e60f8e38c052d68d1231ed
13072 F20101129_AAAFJK fernandezdiaz_j_Page_111.pro
42e8e2521102765eacc3019f73d80a9d
6a68c2cca73957ad417200f7be36542a5bd6ab5c
46982 F20101129_AAAFIV fernandezdiaz_j_Page_015.jpg
1d0dadca08ae2f63ee8c44d7a5eced5f
eae64681970fa2833ff3a80d943a09ec69a15fc5
1761 F20101129_AAAFJL fernandezdiaz_j_Page_066.txt
d8dd566e2940a77e178c906403929413
dc09c501495703dd310b4f3fb3b63a51fdd4045c
1051940 F20101129_AAAFIW fernandezdiaz_j_Page_108.jp2
99030fef392006367dcf5c886baae09a
9eba3e2ca0cb980c36c4c553e6e54d0ff2e10959
114652 F20101129_AAAFKA fernandezdiaz_j_Page_007.pro
60c1d5198305d2109d28c9802a1f39d1
4b58206f643404433e199a53f1304a80bcab251e
24401 F20101129_AAAFJM fernandezdiaz_j_Page_065.pro
3f6407f482ce594736009d76a5cb059b
42df53d38b34865f4ba6178cc99647c89ebbc9f2
F20101129_AAAFIX fernandezdiaz_j_Page_033.tif
5b0a47e9e38797cb0dfcbd7c84a8fccb
ad13d9c6052794d4bb6325c3500354a3cf55e794
21026 F20101129_AAAFKB fernandezdiaz_j_Page_014.QC.jpg
a79ed4c470c1aee6329295436e53b480
ee42dc77af6b63bbee458c1bee59cd4a07d6c657
6840 F20101129_AAAFJN fernandezdiaz_j_Page_067thm.jpg
1d24ed9547261d23070d3a841ca193d2
bd2ddfa15ab9c5b7e0739f239f2da3f778abdbd5
85898 F20101129_AAAFIY fernandezdiaz_j_Page_106.jpg
3c126cf9dc7288b87dfeb993957aaef2
4db94786c6bca47f810385daa953a2568475411a
1051967 F20101129_AAAFKC fernandezdiaz_j_Page_054.jp2
f2db69376031c813f13aa96212bed4e3
2827e9a39f024e2b75cc7c5a1935830e285a07f3
1778 F20101129_AAAFJO fernandezdiaz_j_Page_034.txt
c39b22eff9c822a1ffd48858ee6d9e46
1db3a14c45aff7d1e78409d5fae5b385dcaa5071
113606 F20101129_AAAFIZ fernandezdiaz_j_Page_095.jp2
c1cde15e0dcf7b9ced26d79421ac9b46
bcca4d060f32ff4036017b180e2e2be7f973c8b8
86353 F20101129_AAAFKD fernandezdiaz_j_Page_038.jp2
ed4478b6ba063e0bfe8e610673283586
8d0b88f74cfb56134402a007144ad7081838804a
7907 F20101129_AAAFKE fernandezdiaz_j_Page_130thm.jpg
5dcc1240b8c9f07ccb718cb17c7928bc
eee93073244f76f67df04f46b85ff70b3b35d069
32974 F20101129_AAAFJP fernandezdiaz_j_Page_081.pro
f6c9966faf0d944e18441c25974ff1e0
6768edad0d4007b70c5fd331af4ec0d2b98f29b0
3178 F20101129_AAAFKF fernandezdiaz_j_Page_002.QC.jpg
d1eae95a3b2d2a65635e973d3054a294
69b1aebeb125b7672b642b35e7181cf18a195d0d
54323 F20101129_AAAFJQ fernandezdiaz_j_Page_040.pro
dfa39f700ceea7083bcd70954613d9f1
1da27d5367839636f5e50af51659b7d655262a06
F20101129_AAAFKG fernandezdiaz_j_Page_043.tif
95e696b88bf20b4fdc8b39ba8244a4d5
be83cbe521eb19906c5761dcf96eccfb2bb830bd
836135 F20101129_AAAFJR fernandezdiaz_j_Page_033.jp2
5db1dee58bf8a22c62b45a67d4aa8809
65f51ce62096d04f3fe944340dbde0718bb0102d
74557 F20101129_AAAFKH fernandezdiaz_j_Page_011.pro
b3c4cba1fab0b094136a7a5ec597a074
22a7315b24f5738bb5f07836a9be180a3070b8e9
71251 F20101129_AAAFJS fernandezdiaz_j_Page_080.jpg
b46737df5cfd48ed812353fb0a3eb245
15ffb9709f97d6cf3c945c10862298ec99d719d5
F20101129_AAAFKI fernandezdiaz_j_Page_029.tif
ba141cb1c3cf0597a6a1e9e28188e8d1
a2d548804a2f539fb0b71494f299490947938d5e
79457 F20101129_AAAFJT fernandezdiaz_j_Page_071.jpg
21d5c354220908e08107f2c92b60230a
bfac8df2259d1ebc83eafca6caabe4316dc08d35
2014 F20101129_AAAFKJ fernandezdiaz_j_Page_039.txt
ce988a11ce59684c014f1fb6f71a3238
64cebe774ad3734a1d6f24ea7cab0ff09eaa05f0
F20101129_AAAFJU fernandezdiaz_j_Page_006.tif
b5aaa0f7002db2b3dcf51de0aea5c690
493052b57e3c27968881cc02dd808b109dead554
2433 F20101129_AAAFKK fernandezdiaz_j_Page_105.txt
b927d22a526348c04ec25950866f632a
f0a69dea5f4ef07fc2b7ac52cbc5dd8ff06485d2
29215 F20101129_AAAFJV fernandezdiaz_j_Page_128.QC.jpg
e5d10ec4329983e0831113e1d573652d
5909da4a1ae0c7504e276a1d0c30c4760d96422c
24857 F20101129_AAAFKL fernandezdiaz_j_Page_110.pro
262a44d2bf3f769d0eb6415c85e0ab07
4a9efc64147fdaf25b2dad399e63b8741c69b11d
49068 F20101129_AAAFJW fernandezdiaz_j_Page_039.pro
338ff9469e8d868b3d89a6cbb7c29498
6a3da857dce9284700de94c199d1883d50d72dc3
4552 F20101129_AAAFKM fernandezdiaz_j_Page_007.txt
73454cfeec7cd56d9ba220d47537b810
4ed31e2593e4c814ea229327052e717ca750fd02
24668 F20101129_AAAFJX fernandezdiaz_j_Page_036.QC.jpg
e0bad8867d206d74a8be51d6e0805164
e348930284b5d0074f67a3e8768ea3f42c120633
26745 F20101129_AAAFLA fernandezdiaz_j_Page_043.QC.jpg
145f263e6e2ad855293716a1a013b174
144d32b3d16137924e2bce3cd7d598e11a245605
F20101129_AAAFKN fernandezdiaz_j_Page_065.tif
b96d28a5142d5285e76a976aacf2d058
1f9eb06450d2647a268863629a9be8404daffb74
27124 F20101129_AAAFJY fernandezdiaz_j_Page_114.QC.jpg
3298eac6250dcfe72f91d4e800ffbf99
8b001d21f25452f0c2f8957250b96643cd59e9e0
32549 F20101129_AAAFLB fernandezdiaz_j_Page_108.pro
ee053906a93af41e570e4efe4027a337
0a240a90108d535be900059046f3cad4b6b50ef9
43151 F20101129_AAAFKO fernandezdiaz_j_Page_057.pro
be5f49bbb708d2692366126bc7419142
184ec1b3b23b131238ef448da5b20bb555d9676d
26061 F20101129_AAAFJZ fernandezdiaz_j_Page_046.QC.jpg
b7a9babae876b3e50a93ef2c6c3da4c8
85b275453dbe49432442e5ab352f81dc38b9078d
6831 F20101129_AAAFLC fernandezdiaz_j_Page_018thm.jpg
345afb92fb8e526e8c73a6165832c7a2
5bd93f2382688aaf2d5635a8b9b4099f441f01df
38821 F20101129_AAAFKP fernandezdiaz_j_Page_079.pro
d1071c3b64cc485412e5a6ddece61cc8
c3060ba1e715369d054c4134ede79b76950a52ce
1051882 F20101129_AAAFLD fernandezdiaz_j_Page_104.jp2
916a8b85f39d31cf539064a01c90ae0e
fc0662151f37c6d4e4a9258b6c2447acd2a8c13c
5472 F20101129_AAAFLE fernandezdiaz_j_Page_009thm.jpg
e3e99dc1e51f8be6cee992a920e0f8d2
fa104efd972a2afa58b426ec10b691dd099b7744
F20101129_AAAFKQ fernandezdiaz_j_Page_050.tif
dfdb2804b1d6d8bce9d1a73e30263bef
38f5c0784fafde3053a66a8b3fd74c52a724b612
F20101129_AAAFLF fernandezdiaz_j_Page_008.tif
6b2662c88b80f3412bdbadcbdff43ee8
0a84271094a0f381cff99a614a8051057f3681bb
7180 F20101129_AAAFKR fernandezdiaz_j_Page_132thm.jpg
6c9efff927f51a330181fa270034524c
9998052c6c4dd278b917dcc464328145f44c1828
F20101129_AAAFLG fernandezdiaz_j_Page_044.tif
d12cb51440c347dd7b108b0ae65619df
b171415ce69c99b9012aecfae95d484a281df127
15642 F20101129_AAAFKS fernandezdiaz_j_Page_003.jp2
051fc31608c6661364911b3fda86a1ba
14d60f304afbf92f10cfde71e2303a6b69383cd2
12550 F20101129_AAAFLH fernandezdiaz_j_Page_013.QC.jpg
4e2ffb63875cab8df0937e5da7a40c91
f8cfff6cad905f3cae71c6b045746c52c8126d81
6823 F20101129_AAAFKT fernandezdiaz_j_Page_085thm.jpg
2d5959b58d806f3fb7f661e7974ce35f
c80ced7c97393d22bbe4f8b60e3df972a8d6a9c7
F20101129_AAAFLI fernandezdiaz_j_Page_019.tif
581179289019b62fc7583870f4a1dc77
79b8477ab0da3c3e46e9c423ee9f81664f3e7c83
21502 F20101129_AAAFKU fernandezdiaz_j_Page_109.QC.jpg
70c6f16706cc44e5173be1d2b139629d
5a532e2f6cc6b141b1ddb23bf29d03397e55cb21
114238 F20101129_AAAFLJ fernandezdiaz_j_Page_028.jp2
10ea2955c7d413e272ba4af1336b9c61
43b111d7ae109a96ef2e803491a053d331c98055
97238 F20101129_AAAFKV fernandezdiaz_j_Page_066.jpg
503c229fc2720950c440ea962fff07b6
2685507878a94b080d3839436b198af0b99bc7ee
2055 F20101129_AAAFLK fernandezdiaz_j_Page_083.txt
d227d9e550aa42572b4fc98016a81354
f337a4fec8db8dda02382bbabd29593a11d9c89d
25544 F20101129_AAAFLL fernandezdiaz_j_Page_012.QC.jpg
af311db80d06e38e14bb8a5ec4902ac2
79dbd1b2f042b8cbac0952e0c389cecfc6c0a2eb
25504 F20101129_AAAFKW fernandezdiaz_j_Page_079.QC.jpg
f848e74c1cb2543fd42e2222ddbde75a
8b51af819fafd128ceec066d077714f8c7f778b3
158471 F20101129_AAAFMA UFE0021101_00001.mets FULL
f31a5bfca856ecb5ca14e51477a2e6cc
6283607e711e914c39da0964f66c77be08ca5a96
44491 F20101129_AAAFLM fernandezdiaz_j_Page_070.pro
3cd1c4157530db469e9b60da9967111f
10e1dff282053bafb8627431ec6de439cd4b93ad
52806 F20101129_AAAFKX fernandezdiaz_j_Page_062.pro
4f94062ebddb5a68112719189e940df2
334269baa20a7594dd29ada1314f56a0fb9a819f
29599 F20101129_AAAFLN fernandezdiaz_j_Page_087.pro
a5c4c7653752203748f606cdb9142f2e
50dbac23e20af7a78e24ec51eb7050257706f595
1774 F20101129_AAAFKY fernandezdiaz_j_Page_070.txt
f20b45183936de238bda8e8631cc9381
c438a37c7f8b4de61d074947b48e3d7c42208e1f
74692 F20101129_AAAFLO fernandezdiaz_j_Page_039.jpg
84e2cffc772676a230be1be24da62a6c
448519ccb2cfd227f3a3c78fae6aa8008966a335
800846 F20101129_AAAFKZ fernandezdiaz_j_Page_119.jp2
97d96f97a59af507ea9e169f99a36a66
f1632282c5cd4e57368a6c70aaaa78b515ceb045
23998 F20101129_AAAFMD fernandezdiaz_j_Page_001.jpg
17293b9b154ef0c32fcf242df32e7f62
5aa42366dc87f6b932fe9cd459faaa0860d341b3
43576 F20101129_AAAFLP fernandezdiaz_j_Page_078.pro
e56400e028e303b6bb73cd21b81adfba
8c6722114fbf9451556c32044ae849ca9480d0c6
10352 F20101129_AAAFME fernandezdiaz_j_Page_002.jpg
16b7d688fd933d4ec95c954e2ff1d653
ae495b8ea9cb1311b1c4ee11d2ec374c5bd6591f
65227 F20101129_AAAFLQ fernandezdiaz_j_Page_009.pro
109ff6876062ac25ca2ce90fe4f46ed8
9660faef337fbda1f95237d0f0d68ecfd1dde951
16916 F20101129_AAAFMF fernandezdiaz_j_Page_003.jpg
c70d7265ef235db8de075b755d9f260a
ea3d88cf61b8cdf244093158619876d2ae1d0475
43003 F20101129_AAAFMG fernandezdiaz_j_Page_004.jpg
60e4f43dbf0ee59897a567fd9a871d79
52d5aad3cad9f654d32e3bd6e1387cd859bd3fa7
F20101129_AAAFLR fernandezdiaz_j_Page_002.tif
902ae70ea9da7bbdec90449c14f3d441
d6b37017ab566aaff8dad2a9f8b51064c6ca82a6
83368 F20101129_AAAFMH fernandezdiaz_j_Page_005.jpg
814d7fdfbc918561d921761818dfacd4
379aa79389f198ec4fdf88d2f7df71346802e01c
2814 F20101129_AAAFLS fernandezdiaz_j_Page_126.txt
d02e0cb4d0c15de286e18df5e6c583eb
c366fabacc292fe0ad7ed013a67d41b04dfe18d1
92597 F20101129_AAAFMI fernandezdiaz_j_Page_006.jpg
e48a1da273b78111e3d5737235a04650
7376ae87d3dde25dd781fa0d31595f7c78605eb8
11144 F20101129_AAAFLT fernandezdiaz_j_Page_115.pro
af0a4bc5566866f78096647e164fd3b5
066f2d430ba7d61f35902c0f5d3987636b172ebe
97262 F20101129_AAAFMJ fernandezdiaz_j_Page_007.jpg
1d35ceae21c7cb17517e9c586e41d85c
08b681021623ea90e57a9be1bd4f02b68e5fbcef
F20101129_AAAFLU fernandezdiaz_j_Page_011.jp2
7dae6930aa639d40087723fbb04ed178
f6bbb9f4be2ad81eda72ad8d99e49d1d43f7cd1e
22087 F20101129_AAAFMK fernandezdiaz_j_Page_008.jpg
99374e9e1cfc53cc26d900a27fccebbe
b6c17befc5247b6a7fd68702aaaa9581e65dead1
97741 F20101129_AAAFLV fernandezdiaz_j_Page_107.jpg
19203ad2cff05cf361618419c830ba58
5b37c4c83aa7d39cb1f8427d6c5b30d697a20601
76184 F20101129_AAAFML fernandezdiaz_j_Page_010.jpg
020f6c6383b79d3725eb649ab1788e86
ddebaa6a845ae6380b718871bae949e437c273a4
4892 F20101129_AAAFLW fernandezdiaz_j_Page_121thm.jpg
725668fe0203cccefb18bce08c73a3b4
f9cc5c2d30671bd2c4bb6745d0fdf89baf44b9cd
88461 F20101129_AAAFMM fernandezdiaz_j_Page_012.jpg
11cb20f22baa0106b2e26850df72fdf0
7f8b2ce2485ef1937db170627c72157f2f352b7a
F20101129_AAAFLX fernandezdiaz_j_Page_090.tif
e48d14e543c2f9faeaf4376a790cf429
0065d22c03df112ea9651ae8c40f2acd28691660
72055 F20101129_AAAFNA fernandezdiaz_j_Page_042.jpg
f9733bfd658eb2c41a09c629bd6e4a96
ea229bf4665ff4f4d7213b296dcdc8c4fe4e9695
75995 F20101129_AAAFMN fernandezdiaz_j_Page_016.jpg
2ea589b18106a0b2c3ae1558d3a3b385
c7a318c9a465a1dbfa314825d0a2024efdba7332
F20101129_AAAFLY fernandezdiaz_j_Page_089.tif
587b3ddb4402076030b6b551403040ef
1b642db50ae5efb35e6805cf11e5e5300134e6e9
81365 F20101129_AAAFNB fernandezdiaz_j_Page_043.jpg
1d390419458c571f4f701334a9a2d035
567928eb6149b99d2e9fb6f58e67a204231eae0c
82459 F20101129_AAAFMO fernandezdiaz_j_Page_020.jpg
5cb060a8607adebab0567b5cee5bf16f
000f7849015c758df843060ec0bf60f84c0aa858
7909 F20101129_AAAFLZ fernandezdiaz_j_Page_131.QC.jpg
6570db2818a8d15261ecc22cf25971ae
3c9f23a1019aa0b199ecca128c35b661b5ddd36e
80523 F20101129_AAAFNC fernandezdiaz_j_Page_046.jpg
ff235ad17e31d8506c8440a203688b74
1cb891583c7f819858427b07f74c706185c5fe3d
80230 F20101129_AAAFMP fernandezdiaz_j_Page_022.jpg
202de0efdee21ab808437ccb8a7f2b19
22176da7431623b2b6768130ac9757d8270d7fb0
52927 F20101129_AAAFND fernandezdiaz_j_Page_048.jpg
fbd8aefbb6d72a0751ae5551adaad261
9f158a49dda8cc877ef301d074f159b4585454e0
75594 F20101129_AAAFMQ fernandezdiaz_j_Page_023.jpg
45e00b366f4041de702bd9a37a2f687f
e653459e0bdd8a24514ac21b9d7bfa0aa8e49c01
71466 F20101129_AAAFNE fernandezdiaz_j_Page_049.jpg
6ec98a43e0963a2c94c5f2f2f4243c09
747c4ed5844029427ddab9bffacc55a6b6e1486f
85869 F20101129_AAAFMR fernandezdiaz_j_Page_024.jpg
afa6e38c7aaa9d40c337e598de67e155
f43919c8ac945d6fc9fa41e132f80bce35d76694
65439 F20101129_AAAFNF fernandezdiaz_j_Page_056.jpg
95b59e5f6fbe02aecf9f7cc5e4d153d5
222d0a1dd2704c6ab97faf22f3106527dfebe725
78736 F20101129_AAAFNG fernandezdiaz_j_Page_057.jpg
5a93b2f4bce393b2f6eab581336300dd
f1967e20787483e46a656e31ed0c48c1e3281032
74506 F20101129_AAAFMS fernandezdiaz_j_Page_025.jpg
2fbaddc95dfd7d75483bd16919a164ba
0bf2736fbeba493bbc0c1ca637808b71b27412b3
70956 F20101129_AAAFNH fernandezdiaz_j_Page_058.jpg
d87a71eba8bbaecb3ede7f734cda8d93
3b101738835613a4c96169a09ecddf3c407198b8
83747 F20101129_AAAFMT fernandezdiaz_j_Page_027.jpg
7a2b3db89cbf0509b4318b961aa8e0e1
d99f856118ffd6d40454fdb36a138f17bd9599f7
72353 F20101129_AAAFNI fernandezdiaz_j_Page_061.jpg
a6b1e70eb3fc30e49b312c9c52f99dd9
f3825c3b9ad9b2db3193755720252b86aefd94a4
72731 F20101129_AAAFMU fernandezdiaz_j_Page_029.jpg
0a266e545eb98c8c4e68e2a576fdba31
71bbc849ec3bd2921c641e3af23cc857d2677c17
74487 F20101129_AAAFNJ fernandezdiaz_j_Page_062.jpg
d0254746d27ff93f12f586184aaba405
aa2f9cf95618c33126c1fdc98dd2fd846a0b8bba
77845 F20101129_AAAFMV fernandezdiaz_j_Page_030.jpg
6759169c91e19a2400860eb988add3d1
3dfabd2da9006f9bbaadeef092c963b9c83fac35
89390 F20101129_AAAFNK fernandezdiaz_j_Page_063.jpg
40e324846a66a913175705bbee199b0e
c106e66b8872b8103cb911a0b4760b111c81ee20
57061 F20101129_AAAFMW fernandezdiaz_j_Page_034.jpg
adb3acb783f4435c72b0d65db9ffe50b
f8f3c37b694e2f47a8f3aa902f8be7183dd5c850
83681 F20101129_AAAFNL fernandezdiaz_j_Page_064.jpg
120d37c758d4145d07a6ba35623e9e66
686d6a1d4e8d3b0945a5ee8bb2643fd9e2a4223f
69127 F20101129_AAAFMX fernandezdiaz_j_Page_035.jpg
a55d6a78b31ab763d63f7f869730a357
daa26e8abaabeebed5a492d9f233f379ff649050
70177 F20101129_AAAFOA fernandezdiaz_j_Page_094.jpg
3a0357d9f3650f1dfd275e52335bd2ff
d89d8ed2091555b14a23356666386d1a42b7bb54
81811 F20101129_AAAFNM fernandezdiaz_j_Page_069.jpg
901cd6da8f547a23f5474916a2a91b32
adb014087de4cb479aa097d8cc705f348184ece1
69962 F20101129_AAAFMY fernandezdiaz_j_Page_037.jpg
1151911bd74a02e50948198af53cbb13
88f66b406eec39ce9b944b3f2ac3693e432551a9
79329 F20101129_AAAFOB fernandezdiaz_j_Page_095.jpg
55e10db06c5fac088b983f797a351ad7
288a0e817a5463f26b02e32c0ce78eb005ff9b09
83355 F20101129_AAAFNN fernandezdiaz_j_Page_070.jpg
9f1e22bfd876455838b6288e309f571d
aac8b63d7b4b30f9f06226d5dd56c74d30414f1d
56797 F20101129_AAAFMZ fernandezdiaz_j_Page_038.jpg
3465f3cd4496fac4a1e98a570e68dde5
f0b0a8113f473d1db7203bc193860b2a7239189b
69006 F20101129_AAAFOC fernandezdiaz_j_Page_096.jpg
5b286f304af51ee6b4be3c60cb7f8386
f145fb4717b0bfc7db879343b2594e3fd2a44e95
87375 F20101129_AAAFNO fernandezdiaz_j_Page_072.jpg
8059cc2867257979701987caa2cf0e4d
74a50e7096c5ef5e61377504531bee2888350905
26710 F20101129_AAAFOD fernandezdiaz_j_Page_098.jpg
96fb4649ef8b123fbc64f2fc1c1609c5
4cffc269e9c7f673d9bf9a912ac4f836feb2d34a
38011 F20101129_AAAFNP fernandezdiaz_j_Page_073.jpg
bea62db1f72ab522e727e2ea61984b6f
a023d3ab65ca9d11541dedfab8787a2977f24cbb
80332 F20101129_AAAFOE fernandezdiaz_j_Page_100.jpg
8b7f3ff933ea04a3ec45ceaf4c7df017
4deba43b2cdadc4db69e2469401bfc4651122790
74701 F20101129_AAAFNQ fernandezdiaz_j_Page_074.jpg
31dbb5b35dd1867e2f187b1629ae9c35
529ee328ce5ea92fd5e87d5043a1ccd729355775
78468 F20101129_AAAFOF fernandezdiaz_j_Page_101.jpg
adca045215c2fae8b682f85eb275c4ae
8d465e4b581b72ce16f974d681d0b254675c0ba2
78013 F20101129_AAAFNR fernandezdiaz_j_Page_078.jpg
59f8505d4a583284f9c4e3d9689102a9
8f1eef12493049165dc18b953ca8f60704078e2e
68678 F20101129_AAAFOG fernandezdiaz_j_Page_103.jpg
9f2b3cdd42a3f776cb6185b6bad694bb
b66901507f0709d65409d869e5ef43f86d7bddb7
78132 F20101129_AAAFNS fernandezdiaz_j_Page_079.jpg
1a271b6c2544828036b9758f262f8363
b3451d5fa158890e2a6ef88a460dfa50236c070f
75479 F20101129_AAAFOH fernandezdiaz_j_Page_105.jpg
11c3ba3c5e50dc0352210a377b758357
f49a6c7be176a69b75cdbcb419c18e88fe88eeb2
76549 F20101129_AAAFOI fernandezdiaz_j_Page_108.jpg
3dd3d4382039b6c7590532d150e4396e
311f5d8d9d7c2893bd227a886a2443fe1941975e
70610 F20101129_AAAFNT fernandezdiaz_j_Page_082.jpg
1eef1a6686b50770555a52739c0810e1
8ebe26bd80d50c430d3676aef42bcd30a1a12d60
68960 F20101129_AAAFOJ fernandezdiaz_j_Page_109.jpg
2fd63df501eca760ceb9ca2186db6515
3de0c94f2024784cf16079543b49bf7fe8089977
77583 F20101129_AAAFNU fernandezdiaz_j_Page_084.jpg
4fa4b711e25f46fb9a6c1d12a41ce13a
00318195d5a2d76df1015fda613dc98bfa47019a
62733 F20101129_AAAFOK fernandezdiaz_j_Page_112.jpg
8a01bceef0b450e8a0f846a7ea1a8ecf
64f3326eddf2d6429b77b6b8bf947ad7537f9a3f
75234 F20101129_AAAFNV fernandezdiaz_j_Page_085.jpg
fb7b878d4b5bf06d6ecf1f26ce2469fc
01520be029b7e442fc363255699cff210a73b133
73447 F20101129_AAAFOL fernandezdiaz_j_Page_116.jpg
3cfcfda17b201462ff56675e40a2b99f
4f72bb5973651c8794b1fa2dd002adef6d58d213
76448 F20101129_AAAFNW fernandezdiaz_j_Page_090.jpg
3b981c4a5c9a592d65feb1a2c82e8df3
161f81a52ef8461edae4b61d12b00e92b246e1a6
1051976 F20101129_AAAFPA fernandezdiaz_j_Page_005.jp2
70a625fa0482a5a26d8929aeb6619138
f1ccb4b1e1738b00239e39463bf8d059e3f07d37
75030 F20101129_AAAFOM fernandezdiaz_j_Page_117.jpg
fe2d645ece6b24612abce698dffc81da
28d9e828071ae60511a5a0b511178ae204c6323c
77425 F20101129_AAAFNX fernandezdiaz_j_Page_091.jpg
4a1d2b7994cf9b7dbb942be0e6fb5bc8
d0caf95faee0de2b8e089f7e4e0b1fe6f446f65b
402243 F20101129_AAAFPB fernandezdiaz_j_Page_008.jp2
d4a7b526ba2063d26dd99607648798c7
d0302075c63b46984b13a1be52782041fbf7bb1f
71924 F20101129_AAAFON fernandezdiaz_j_Page_118.jpg
dc0036e0ec0e2eb98ed79848aa20e72e
203e6aa7b67a4d567ae7affef21cc257fc5c433e
64824 F20101129_AAAFNY fernandezdiaz_j_Page_092.jpg
c6816023f7f063fa5959fa19ea06bc16
a05923d906f28048a14f1ec80958d9b9a71f8f2f
F20101129_AAAFPC fernandezdiaz_j_Page_009.jp2
d4553f131b82fb3d53a018909a2918b1
6a903a391d3d8fdef4e25bd37f54af56bbbfd964
51843 F20101129_AAAFOO fernandezdiaz_j_Page_119.jpg
04370f3559554667607e2c0b9412f0d4
928e0f675fc27f2e7ae513a7fd76981c390209dd
58542 F20101129_AAAFNZ fernandezdiaz_j_Page_093.jpg
e9e41fbc3fc69e93c072bc337a615642
034cea4c74609dbb71d1b9c1820bf5dbcd2e71a0
1051965 F20101129_AAAFPD fernandezdiaz_j_Page_013.jp2
3fe784d41188852faf86b011c78f4543
a55a247d213e80134f8165c94a28db284208e65f
47576 F20101129_AAAFOP fernandezdiaz_j_Page_122.jpg
553e0c734f6d5f53636fbd0a837bcc46
a410654db9c7db638af8f81a92c6a75731fb49f8
99225 F20101129_AAAFPE fernandezdiaz_j_Page_014.jp2
0fec70e3d465658be3a0822b2b99a911
11a6eaa1c1fd130663d721b5fbd71849d57ced92
47958 F20101129_AAAFOQ fernandezdiaz_j_Page_123.jpg
08dc4a83fa1cd4aa157e7976191cb041
f30d10143ef02e8eb381d33203e17c39c3ef7789
115667 F20101129_AAAFPF fernandezdiaz_j_Page_018.jp2
e183817612f369ba0717cd85f39366b9
2c83079d53970ce7a810da84e2067032b15069cf
108132 F20101129_AAAFOR fernandezdiaz_j_Page_126.jpg
12b5269b8afa376fdec3dff8c94bae84
c3d1854f2e094fbccd1bc706727201f45f1cb9e5
115502 F20101129_AAAFPG fernandezdiaz_j_Page_019.jp2
97076a079b735ffc4931e202e857d554
40df4a1598c2a753939c22ccdad2596e2dd51ab3
108421 F20101129_AAAFOS fernandezdiaz_j_Page_127.jpg
29490c3f62411d9d03b6d12e1be3ce51
df5d4a25ccbc74b20f323fee5f3a05c9d28ab3ae
F20101129_AAAFPH fernandezdiaz_j_Page_020.jp2
7257f8a14a7aa35557f52691a0ede2ac
a0450e2b02a1047d78a6a45dc747b9bcf9990c96
104493 F20101129_AAAFOT fernandezdiaz_j_Page_128.jpg
d9865ec2b1a52bd68897ab17cd440901
29aace75c450eb3042e924004b93f5acd2e6e0e3
119995 F20101129_AAAFPI fernandezdiaz_j_Page_022.jp2
e919ac0cb46a67b1b3115b94d0dde33e
98eba46799513054a5c7b7daa95ef0fe34208eeb
1051950 F20101129_AAAFPJ fernandezdiaz_j_Page_024.jp2
ed777a43b898a9915026f8778f27c493
29363738b270d4b8b8415dab37965e54a57c7f32
113790 F20101129_AAAFOU fernandezdiaz_j_Page_129.jpg
a73dc1c468505336639257421d70e3db
9bbb7565990997dc12d8e288d72f8eaec6a84314
1051979 F20101129_AAAFPK fernandezdiaz_j_Page_025.jp2
dd12f12bc131e5ad0f0732c8ce3b5ad6
01e1b95fcd6feff2ead2f338030712ad8cc2cd91
114883 F20101129_AAAFOV fernandezdiaz_j_Page_130.jpg
59e056117572e7ebda598625154f70ea
a6897e72056dfe8700a894458a47c3494730fb77
1051984 F20101129_AAAFPL fernandezdiaz_j_Page_027.jp2
5b64061bc4bca323494d47eb5f153352
1f3a66036fca8a5ba412cd80bb5c44ab24f6554e
25583 F20101129_AAAFOW fernandezdiaz_j_Page_131.jpg
89af901c70a72b48f06870d773d3d415
675051120af7f590c24bd0e76b72adb9b8cabb02
109239 F20101129_AAAFPM fernandezdiaz_j_Page_029.jp2
ff8d995828ce712e7c203869b335aec2
8cb06d2c1650c9d96de451935e9475c1de5eceda
80911 F20101129_AAAFOX fernandezdiaz_j_Page_132.jpg
a9a5ee09fcb73afbe25ced4575088c84
f22e0d5286db9321609098c57780391c1c061197
1051971 F20101129_AAAFQA fernandezdiaz_j_Page_052.jp2
e54811b98f9150cb020bb88e25d8734f
b46163f627aa8f641561eb17d6bbd3eb91420450
118641 F20101129_AAAFPN fernandezdiaz_j_Page_030.jp2
c2e588302864af8eb91fde3af165adb5
2cc6d4bf9ae332ee6bbba4da2a5d75da6f2b8f54
5962 F20101129_AAAFOY fernandezdiaz_j_Page_002.jp2
47a6c2f121185858d12bc7a8b37a937c
5499a77b8df8e9ab225095a9af8471a65a6aff7d
116112 F20101129_AAAFQB fernandezdiaz_j_Page_053.jp2
3cad0d9f04fa17397ff7c4dc834877a3
1b833e6f5a364813504e950e51f2171b7a38ac42
816771 F20101129_AAAFPO fernandezdiaz_j_Page_034.jp2
8264568e61d598ff1982043e178d983c
f74ab5c6395b7c7d9058a93a4decf6c3ada8fb34
59804 F20101129_AAAFOZ fernandezdiaz_j_Page_004.jp2
e2f6edd818814b1fdca4a1b2bdb44d82
6f09e2bf034ea28f388a5337e07da04e526bfbcf
1051951 F20101129_AAAFQC fernandezdiaz_j_Page_055.jp2
db26f6b5cddfbb6e7199b16fa18b6b75
5b2401aaeb40afb9cc2e32b31adf99f72d1e599f
104277 F20101129_AAAFPP fernandezdiaz_j_Page_035.jp2
3eeca0248db1ed84265c0db5966b9c40
7606a046ebe3a766e20a53a24be38dc587c3066d
97779 F20101129_AAAFQD fernandezdiaz_j_Page_056.jp2
99398642b96fa62aef2ae42fb854549a
e2e518036190e9add854c73d86dc28d51fc2aecf
102578 F20101129_AAAFPQ fernandezdiaz_j_Page_037.jp2
a65f8057539030dbdded3fd358aaf074
3ee55e22da19c6441e998eba56a2328a4d7e95fd
1051841 F20101129_AAAFQE fernandezdiaz_j_Page_057.jp2
85a688bd7be015763a1da8812a933b1e
923b8f591dae791db34b2ea0b46e8ced7ef877d9
108482 F20101129_AAAFPR fernandezdiaz_j_Page_039.jp2
04a4eaec4e43e947a46e89e052399416
8726690701078b30a26dd85046afb84bfb2b0fc1
104015 F20101129_AAAFQF fernandezdiaz_j_Page_058.jp2
a4671192973cedf7a3df60e344983e69
39f742ab6521b7f6342a70b54ae4b098aa713b2f
119906 F20101129_AAAFPS fernandezdiaz_j_Page_040.jp2
a568279e3c64f04b7b1ed5e841d636bb
2de67130005e26dcd7df1482cf21d30e1e92bb5d
1051966 F20101129_AAAFQG fernandezdiaz_j_Page_060.jp2
58d8621f5421ccad6d502a3d1fd7bdcd
ee46264f86a3ee2c9b7ca6f31a125b30cbef05bf
1032078 F20101129_AAAFPT fernandezdiaz_j_Page_041.jp2
6ac3706b8414717b29dc29fb291e9936
b556bec4f134a0955837e88a956c29b021036316
1051972 F20101129_AAAFQH fernandezdiaz_j_Page_063.jp2
42aae9d60db9df77ae71f5600f36bf01
3339a40fd0618d2e01d3e3ec0b9496fe2ac86c93
F20101129_AAAFQI fernandezdiaz_j_Page_064.jp2
e2e7b19ac3026bfd5585e67b3e5a464d
aaa8aa0c1c3c98794d9622584123c6b574f26e43
121881 F20101129_AAAFPU fernandezdiaz_j_Page_043.jp2
ee74572f6da4a77feddf576b0c4eb1d0
7fe01d4cafa1c65e7711f85534c99350452af595
111289 F20101129_AAAFQJ fernandezdiaz_j_Page_074.jp2
2b953e900eb08352e5b30117eca32ece
72a1f0ea1cabbccfbe2240842d62c93d2dbe9fd8
F20101129_AAAFQK fernandezdiaz_j_Page_077.jp2
82712bb78cab9c0787c807dc13319dd4
e25ba9f8ce1a30d57198c4b620b853be9b4a1a51
1051819 F20101129_AAAFPV fernandezdiaz_j_Page_044.jp2
ffa47e9cf8da2249dfdbd01265c143bf
8445f7ac4333ab4f9fb3034c6595cfb9c117fccd
107385 F20101129_AAAFQL fernandezdiaz_j_Page_080.jp2
15292638208711a321c7338387320e96
1f89d7349b01b27a7ff80644aa363352e1b0460e
1051983 F20101129_AAAFPW fernandezdiaz_j_Page_046.jp2
d4fc6529d8f69e040428cde4b0065ddc
799e2658a36438389cd822a7ba14fd3242b347b1
1005811 F20101129_AAAFRA fernandezdiaz_j_Page_112.jp2
c00581dfe9bb9a4c1bc8b5b64b212280
31affe17cffe87ab171507a09ada23cbfe4ffa72
1051941 F20101129_AAAFQM fernandezdiaz_j_Page_083.jp2
f8c639459f17454e93a8c9b489f916c1
a67732d7ee3b2c56cc1538d303e3c935e9410b89
1051968 F20101129_AAAFPX fernandezdiaz_j_Page_048.jp2
30505d41d6e01e7d5cefe9585819e1c3
b5d4c8a66bcd3ac9dcb7d4b0add4ae46ad8db894
93316 F20101129_AAAFRB fernandezdiaz_j_Page_113.jp2
4b8ea570db6243c1acb5d693c44181c2
af27853a0718b4c2be04562ad6cb95272da619bb
113718 F20101129_AAAFQN fernandezdiaz_j_Page_085.jp2
455c8c653dc11593c36d350fdd29ba93
7a2ca4ffaccf3b4d9c8057a6201f59dc93c93cfc
111910 F20101129_AAAFPY fernandezdiaz_j_Page_050.jp2
f6bb1883ad39ae627d0e0adbc1a58e5b
f5c34d0eb564003c73ed870e968ba0773ce14a99
27063 F20101129_AAAFRC fernandezdiaz_j_Page_115.jp2
e0764dd751be1b761b9ff79e043cc2ec
e0b314e530b23b0760083b35f27baa776be0b2e4
94827 F20101129_AAAFQO fernandezdiaz_j_Page_086.jp2
572c82aacc08f4b73df5b21e7b16f440
35f2527ee7a64d8b5b311063855afaadfeeea088
F20101129_AAAFPZ fernandezdiaz_j_Page_051.jp2
fb1bae7264ea86a6dcbdedbb5c417c72
9185e2a85ec8f478bc919e3ba57cc3b69e9bbe02
107086 F20101129_AAAFRD fernandezdiaz_j_Page_118.jp2
3f9b79b22f2823c78940fdc43d33f050
a4655887a3e232e04db47e0c598fb187ac505102
1051927 F20101129_AAAFQP fernandezdiaz_j_Page_094.jp2
30c519cc88f87e73f4a87bb4387aa10c
81d0134ff5b0d10d176cd7474c2100633719cd71
778253 F20101129_AAAFRE fernandezdiaz_j_Page_120.jp2
6266f4984e95b3eb46d5e5e3bb5a4cee
c014ea0826cf9cfece27fd124f2d64ad0a8d4533
1051868 F20101129_AAAFQQ fernandezdiaz_j_Page_096.jp2
7f6502d70d045dba08dbc4916e6fe44f
8f9faebd4438577b06c1c1bbdad793e5c95d7fdf
756229 F20101129_AAAFRF fernandezdiaz_j_Page_122.jp2
5a080216dc616f46ac06b7f040cd3b6b
0b974a6a9f4e2f386013e7a8ad7fc66e7da179c1
F20101129_AAAFQR fernandezdiaz_j_Page_097.jp2
3b5850aa36e6d6f00abd692164715783
6e453ff3d04e729abf88594f528d13241dcbcb45
98401 F20101129_AAAFRG fernandezdiaz_j_Page_125.jp2
ab132973ffe88c1318fff71e2482aafd
87228f968a38acf250c9fc83ace0bd153a3f8d16
33957 F20101129_AAAFQS fernandezdiaz_j_Page_098.jp2
3081f0c82f57907cffb82a7f68433b51
69c1e53914be64a96ffabbc55cf8ad4f1dd10627
1051975 F20101129_AAAFRH fernandezdiaz_j_Page_126.jp2
37c38dd6d3036f9715b2412aaaa90a1a
daa8b8dc1d50c4e9a800e253ce04ac0322579dab
121372 F20101129_AAAFQT fernandezdiaz_j_Page_100.jp2
222ffe5a837c28f626c4a1e5805a45a5
ffc4c86e892e15524e02d7a7939fcacb15905449
F20101129_AAAFRI fernandezdiaz_j_Page_128.jp2
9ed3b153ead00dd5a83fbeeec566a2f2
927acfb322e886db880b055c0c1cc6f89fdfd398
1051911 F20101129_AAAFQU fernandezdiaz_j_Page_103.jp2
6e7902f4e6ff1360435405b09666c924
5a7337ecd6ed895bcf9838c1e728a2fa6d552eaf
31123 F20101129_AAAFRJ fernandezdiaz_j_Page_131.jp2
38151651489453894e595b5a0e24ec80
ac5d194c67b45c39e3008e7d282beb2fdb28d2a7
113887 F20101129_AAAFQV fernandezdiaz_j_Page_105.jp2
6ecbae1d1dd59d2b0314825dfa1f9fa7
ecc0674dd550a53d38a2f0e6c58a7877a3fe9e26
F20101129_AAAFRK fernandezdiaz_j_Page_003.tif
0a40295342ad579b669d05a24336cfa3
9039b06c169c27533b4e29fbcfde7e6396de45cc
F20101129_AAAFRL fernandezdiaz_j_Page_004.tif
1ca5b5c55940ec28c0a20917b6ba96c7
20031ef2d82710dcc55a019d1588bf28fa3dac04
F20101129_AAAFQW fernandezdiaz_j_Page_106.jp2
973b3743e503c7722221b79bfe90e830
55a9d7bbab701efd7468813397463414f20b84fb
F20101129_AAAFRM fernandezdiaz_j_Page_005.tif
d8d7739b19145ffe33736625c061f752
fc9f657cb39a7feb973ee9423701d350b6ce96cb
F20101129_AAAFQX fernandezdiaz_j_Page_107.jp2
76a5d26a08a564fc22db3ab00297e3c1
3a3ca9393c0e3867bd6c0851a364f8e38ded3147
F20101129_AAAFSA fernandezdiaz_j_Page_024.tif
b51b34af881beaf591c884e2fa08bbb1
b9b0591d510f3214649355cad2d0ba54cdc45887
F20101129_AAAFRN fernandezdiaz_j_Page_007.tif
6e4cf89fa33471f0b7a17598aa7816fa
6faedc776ca99baf1eb9c8bb1e86a0a03c1e4788
1051924 F20101129_AAAFQY fernandezdiaz_j_Page_109.jp2
2cbc7cd63bab4d8439984c3640542d15
60d108d48bffea6f30714b2686e4ba78c44d10e3
F20101129_AAAFSB fernandezdiaz_j_Page_026.tif
ef6bc398feb1d7a81fc259c478738c5a
c9096bb7f066da028bf389b139313f80341cbb2f
F20101129_AAAFRO fernandezdiaz_j_Page_009.tif
7a2a8a4c0ae7b7c10399ba87151d41b3
bf6137e4fe1f03c16645030345ed255c3a771591
1051962 F20101129_AAAFQZ fernandezdiaz_j_Page_110.jp2
797f694705b2338874781da93e47e61f
883ee9ca78fd940892caa5bb54a9f276ddda8b2b
F20101129_AAAFSC fernandezdiaz_j_Page_027.tif
d5331c9867d69bad828ce211fa17591a
a6f6177743d97a1f680325484a777dadcb450840
F20101129_AAAFRP fernandezdiaz_j_Page_010.tif
1cb93eb8cc976353d05beb20f6a52e5a
07e7a99512538fecae799a94f88e81ff853d209f
F20101129_AAAFSD fernandezdiaz_j_Page_028.tif
cea32881a4c31266f1b5a5c3d4439fee
6a6649abd4fe5363cad3ef9629043cd27b1a1690
F20101129_AAAFRQ fernandezdiaz_j_Page_011.tif
337bac0257b55b16de287bc603fc4309
4866a80dbfc873fced1a014849c3011022cb908f
F20101129_AAAFSE fernandezdiaz_j_Page_030.tif
a80bee85871b059a10213efae3f3f990
f4e910268e5516e92d4943c9ab03763e5898c1c4
F20101129_AAAFRR fernandezdiaz_j_Page_013.tif
a2ebaca6573e9f1e8737b815d8990d04
93fb8c68b503fe4685a14196a92b528140463dd7
F20101129_AAAFSF fernandezdiaz_j_Page_031.tif
c1c07dd62caa667d8cfc77595a045c09
b3d139c2baf8fc65dd74941d68d6c3c658382b0e
F20101129_AAAFRS fernandezdiaz_j_Page_014.tif
c6d63a4249015cb27e248721762d17d7
2d2cafdac8f2ea9f572cae7f2fa121f87e88f689
F20101129_AAAFSG fernandezdiaz_j_Page_035.tif
20dbef7b3ae3cf6622b84a90ca4d190a
9dfa2dc09e1c85828c30d4f1867444636d198de7
F20101129_AAAFRT fernandezdiaz_j_Page_015.tif
eb4f0f1b37669fbb82ac85f11fee0f4b
2b16df33fe570f2b22d271182b5f290e1d01909c
F20101129_AAAFSH fernandezdiaz_j_Page_038.tif
4f0e2b85a6c2923bcd7dc4f37cdc60ea
7e59330bf94d1c9dce2de511f47549f415b2e11d
F20101129_AAAFRU fernandezdiaz_j_Page_016.tif
4a7e5138d9cc8786e1c1b25bb2ee9966
bd89d71120fd3d835d0eb2b17ae87089a19f15dd
F20101129_AAAFSI fernandezdiaz_j_Page_039.tif
76176db6fd070f57740207a24685e1ad
64012adfdea7222332d2513edd96128b8eabf9de
F20101129_AAAFRV fernandezdiaz_j_Page_018.tif
65a060eb9d963451fe8d113492dfcfde
573cfcd9864d7945212e7c6f7513db75e02bbce2
F20101129_AAAFSJ fernandezdiaz_j_Page_046.tif
e5a28f1001d99d06c44f6738647d1d6c
99452bab2efae885bb63871ecf9ea7e549adb6ce
F20101129_AAAFRW fernandezdiaz_j_Page_020.tif
fffb28beadbedff08944480fc7b1565f
3b4f91c41e60810a294d47def647a3017dad2b55
F20101129_AAAFSK fernandezdiaz_j_Page_047.tif
525ed4bf229d4b7f0821ef3dd0a6ca7f
d46975069ebdbcff16f5c4a8f02722c1db6a105c
F20101129_AAAFSL fernandezdiaz_j_Page_049.tif
2cf2c3d1c852701b32c7c0a616792437
b18605d8621575092e5bfb78384614c72535a71f
F20101129_AAAFRX fernandezdiaz_j_Page_021.tif
9b1f4dcb367427205024724e1eea6e92
3b41e79ebe244cb13ee57418dcc069dc31a66583
F20101129_AAAFTA fernandezdiaz_j_Page_075.tif
a579c58ae3062c1db11b26ea1600cc15
8df86a3ff6633b7b8a54b4bd1e45d85c073de38e
F20101129_AAAFSM fernandezdiaz_j_Page_051.tif
eeaf3e8dee4fa4790ba5053eb1c92f60
8b530db828174e3cdbfc35035807c9e05dd9f165
F20101129_AAAFRY fernandezdiaz_j_Page_022.tif
7e313edf606fac6e2089bae98d9b0951
55e366eae63af96e35a28387a749626a107942f9
F20101129_AAAFTB fernandezdiaz_j_Page_077.tif
3a03f9e7f1b1a42152e8dac0a1ea9f63
92a78560f6f0a61f44388a8498ef09759362141c
F20101129_AAAFSN fernandezdiaz_j_Page_052.tif
0c11c4972641c30c65d8c6025a415600
1b0de1a763c1d72561e797d7163e52e5fec22ca5
F20101129_AAAFRZ fernandezdiaz_j_Page_023.tif
ebc7ac77f09115a44b850bb8c1d3789b
f7790a2c57415ad7e78c5bb867e03bc782789d58
F20101129_AAAFTC fernandezdiaz_j_Page_078.tif
f1d9449d573200a70badbcbc9f7fdcf9
cac7af73c2397752afa4f1acfb1212f84f38591e
F20101129_AAAFSO fernandezdiaz_j_Page_055.tif
83499effad3cd5e7a63c278811807256
ca3efa1c796b0bed74684214e336ae99fe6adae9
F20101129_AAAFTD fernandezdiaz_j_Page_079.tif
fc77411a28e9f4ab400fffedffbba519
03b033b2affb219cafdaefefcae545a58982027a
F20101129_AAAFSP fernandezdiaz_j_Page_057.tif
558d801d1fa719f5a68948bf6d0db588
6800162c14d1a233ca14ea7e968a956b5bc4fc29
F20101129_AAAFTE fernandezdiaz_j_Page_081.tif
f7ad671b956a75baf1c80166bb0fd834
b1ed05eab062687c1c847e55224cd545f9880ebb
F20101129_AAAFSQ fernandezdiaz_j_Page_059.tif
59643130ed381230562fa3606b91163e
74f626c8794507a4355cf506c9d70784dd41a803
F20101129_AAAFTF fernandezdiaz_j_Page_083.tif
f0be0a36fc6d0aaf7efe46f32997c558
57b27754d0f571b550ae682350ca45bff6274a4b
F20101129_AAAFSR fernandezdiaz_j_Page_060.tif
0c39d5576c9551f9f451f649353bc86e
a0e879e4fdbf84382d72cb3c2c41c77538121491
F20101129_AAAFTG fernandezdiaz_j_Page_084.tif
81dc02d726eb4800002132737c1cc898
0045a96d23069c47c7a716fead34d459cef7944e
F20101129_AAAFSS fernandezdiaz_j_Page_063.tif
1ee762d915ccd9abb9bddc527540bd42
7ec58e56f6f3712c08d80c728c4824253ec36228
F20101129_AAAFTH fernandezdiaz_j_Page_086.tif
39e846ccfcd46c7fa625425f4411420a
cc7142eeeccb1c032bb6004010b2760569e5d163
F20101129_AAAFST fernandezdiaz_j_Page_064.tif
e610877d03101d207bcc8f5d082ee9b5
ae01e352f0e7c28f8a022fa32bc35f29aa98fe29
F20101129_AAAFTI fernandezdiaz_j_Page_087.tif
e896a51ced9464289820ebaac4d23044
c077fe762eea1338576a9278a7fea59a88e0547c
F20101129_AAAFSU fernandezdiaz_j_Page_066.tif
8f645a1d23f02f6c7f58942894533c5c
207af86f30e4a17d3645dc5a248370b064ac86dd
F20101129_AAAFTJ fernandezdiaz_j_Page_088.tif
f660746b3023f76c7da40ad00049dc5e
26f46a6b630c93fb154fba45a98bbd6c6a0ac9fa
F20101129_AAAFSV fernandezdiaz_j_Page_068.tif
b5ef0203ca8386e4e34824595f424a79
53ef6782f053fd979a82f2d2d5e130af0c977a79
F20101129_AAAFTK fernandezdiaz_j_Page_092.tif
d31485d8c8edbc97eda86c47555a21e4
225cb392192f04b0b4a6ab1dacfbb16282850d04
F20101129_AAAFSW fernandezdiaz_j_Page_069.tif
82882667ebe62679e1bb5448204a52b2
8f196fb6483e8c0773e6ade999840948d755c4e1
F20101129_AAAFTL fernandezdiaz_j_Page_093.tif
c53921152ae8c386b7b7980e5ee6cd02
06e622ff6de5275e2b45d0a9d6e39dd87e165c80
F20101129_AAAFSX fernandezdiaz_j_Page_070.tif
3dc54cbd7a35738d329a38fe7fe5e4e3
cfeb070f2fac104305f3cbe4adbe53427c455785
F20101129_AAAFUA fernandezdiaz_j_Page_117.tif
652772f1ab99fd94db63484de5f34e20
67ead78c69fa185dd4dbc08b44706221c003f34f
F20101129_AAAFTM fernandezdiaz_j_Page_095.tif
d27eedaefb70967b26f9d1d0075f8eb7
9f8df2298fccf92d03fc0ef453f2fb163bc61539
F20101129_AAAFUB fernandezdiaz_j_Page_119.tif
cc26effb9938be6ceebbac505581759d
74d9bdd2e9df7c7209be8cf3084670d31761fffb
F20101129_AAAFTN fernandezdiaz_j_Page_096.tif
87e5d44bbf923eb4180e13e66fb9eda3
9fbc25363775b9a0cdb0f4d57bcc66beb29a7907
F20101129_AAAFSY fernandezdiaz_j_Page_071.tif
777e397c53b5cad7f1dda19e6f9da732
06c123bd69f708cedbed56c1d986d8584ed3a4d1
F20101129_AAAFUC fernandezdiaz_j_Page_120.tif
3b610580cb31cedb18873cebba1c5f07
1a402739b88c4ed6c502220f934c2b10b65463d9
F20101129_AAAFTO fernandezdiaz_j_Page_097.tif
1434cafa032627fb72238e3fe6e917b3
17012e64d93a4df6dc3f43ff2055e67107f61f43
F20101129_AAAFSZ fernandezdiaz_j_Page_073.tif
619f38ab7e5df8c1988a525ac9f4d525
56dd5f9db503c63b0eba708b316b3e1056868cc0
F20101129_AAAFUD fernandezdiaz_j_Page_124.tif
e23ae7402c25ce6bdaffcbe5d8fa8803
7b1254fd43040dbc705d3db6ce24fba2646b5f9d
F20101129_AAAFTP fernandezdiaz_j_Page_099.tif
8a0cf2bc30312a6c8c2c28bba32acbee
d7f5394a83f2a379a2b0855317b4576bbc280d72
F20101129_AAAFUE fernandezdiaz_j_Page_125.tif
3bfaa9653406b3680a3cf0a1561c96f3
8684627a60d309c284fde363220f9519d10562f4
F20101129_AAAFTQ fernandezdiaz_j_Page_100.tif
5b0138bb765935cda81db9f7b9d9b469
b2984ee38e21034be1a02764eb91be47d0e8f6f3
F20101129_AAAFUF fernandezdiaz_j_Page_126.tif
83558ea12afd85a97711409318984644
d6d7db54df4185043bd811ca12ea836c9fce8c22
F20101129_AAAFTR fernandezdiaz_j_Page_102.tif
e3a273fd96342635063aed68971817ec
15f1ce3493c0285500608c9a7d0a7b8d427bf218
2818 F20101129_AAAGAA fernandezdiaz_j_Page_127.txt
a58e1081c47ff0ef107c6f1319119512
09892ec2f13979f43682a58fd379216481e1f9dd
F20101129_AAAFUG fernandezdiaz_j_Page_127.tif
cc22757e8ff1408f6f649cc790a5d8e3
3825064312c2f8d5004deb335151b96be7bf0bcd
F20101129_AAAFTS fernandezdiaz_j_Page_103.tif
4f5dbe5b7980786d3b8eabd353d631f1
8d85fbb0ede71ca1ea93b32f0fdf6a5e4acf0615
2770 F20101129_AAAGAB fernandezdiaz_j_Page_129.txt
78a3648973306232d1515286cd36d490
a4b76b264ccab56f43abe4c10069c45dc967953f
F20101129_AAAFUH fernandezdiaz_j_Page_128.tif
bb660dde885c987af9c457fa9d46723a
3388099a941a1510d64d77a576aec7c5d04f8efd
F20101129_AAAFTT fernandezdiaz_j_Page_104.tif
38ccdd543a2f54b098e9f6b0d2f84dcf
bb7197c52a7853a857d7c201eb8b091f907bb885
2825 F20101129_AAAGAC fernandezdiaz_j_Page_130.txt
3e281b1c311d0131ec2140102e116d14
550d416d48592f424ecd772bd46704975cd9af69
F20101129_AAAFUI fernandezdiaz_j_Page_129.tif
acf4b863cc181bcc3489ba375831d038
4d4fb1e03cb7c27735a83cb9f457bd0688693782
F20101129_AAAFTU fernandezdiaz_j_Page_107.tif
25e361d4d32f5d376d705a85db47bc91
c90a51b39dc8da92caa347ed7f41213465bc3549
757 F20101129_AAAGAD fernandezdiaz_j_Page_133.txt
47faede705049a940e61296fdef12593
3e926e3d205d3cc88b1d019642cd90ac0926c9a2
7050 F20101129_AAAGAE fernandezdiaz_j_Page_049thm.jpg
91003fd680e59e5bf4c9baa93fadd8f1
6d2c8d0ddb43c82b6b56c67c9092bb482b2f3b2e
F20101129_AAAFUJ fernandezdiaz_j_Page_130.tif
50ff0bd169ed119e6445ad9452a8ce30
fdc2d9f5158b23f035383fb58944a1f7c3baa5a0
F20101129_AAAFTV fernandezdiaz_j_Page_108.tif
db2f0f6a79e3b5a761f984144bd9b7bf
61ea626d0878c675aab6b75b09d3c3d571692c93
F20101129_AAAFUK fernandezdiaz_j_Page_131.tif
4cfb71ce2c189b50b23e9f9f1e64d9fa
1285772038bc090471c86cfab969313db52fd6b1
F20101129_AAAFTW fernandezdiaz_j_Page_109.tif
e8834bb6c2d47a4a334f9630604d323a
ea0f8d52b566b50bce844b5030e6d26f71a8b915
7302 F20101129_AAAGAF fernandezdiaz_j_Page_051thm.jpg
b245e1002f27eedcf4ab43ed8ce720de
e937328c79ba0b73e085be008b41e633271c4e58
F20101129_AAAFUL fernandezdiaz_j_Page_132.tif
421b31e2b23f8bf9e088e093211f6455
004c911efc836196fc6993d561e5be4c085e0a0e
F20101129_AAAFTX fernandezdiaz_j_Page_112.tif
c08dba66620d0cd0d822b52f949e70a0
2b9b3ed8fdd3ee1f8471d88f94160c2c68e54bea
20366 F20101129_AAAGAG fernandezdiaz_j_Page_125.QC.jpg
01b199bc9199ad44683c26d5a48ec05d
236f23849997d7e83f530e8dc287fcee3497bb74
F20101129_AAAFUM fernandezdiaz_j_Page_133.tif
8c8fed029d1da7d23e95fba10502051b
9627b5b5c11eac8466a605269c7ad9e5b3818b88
F20101129_AAAFTY fernandezdiaz_j_Page_113.tif
a55067de5d08235e7c3af1383fb124e3
777141536e7cd116fcb8d21a81df4e881e3ae3aa
47097 F20101129_AAAFVA fernandezdiaz_j_Page_037.pro
a8fb6e543b607f776c539ff0c5cefb29
1c2bc2fa1177d2fc7f65960d8485633f04e31b3e
8066 F20101129_AAAGAH fernandezdiaz_j_Page_126thm.jpg
abcc59b381a6ca9804b91e102a7f5a85
161ea03d44d0188f4d2950857919e957cac4cfe5
25539 F20101129_AAAFUN fernandezdiaz_j_Page_004.pro
8e10ba7b6fb21de7f5eeb690c3e0822f
49d4a6d15a43d19fb8ab810c2cc17a40aa75d511
38495 F20101129_AAAFVB fernandezdiaz_j_Page_038.pro
1da1744804106962422119bbdafd9ca8
3ffc2cb777a411a17ae9d6ba9710a36888d61143
24480 F20101129_AAAGAI fernandezdiaz_j_Page_047.QC.jpg
f9a1bbbb246397608fc5da77ec52a122
f2496d9f68dff0ce969637bd2618a07074f019af
72786 F20101129_AAAFUO fernandezdiaz_j_Page_010.pro
b87f45fe886bf595cd22e604c5ce0920
03ef37cdcd29d529d1ae078152a687a20896c825
F20101129_AAAFTZ fernandezdiaz_j_Page_114.tif
6f0402960422d22bc885316cd6576e03
d51fcd859c9112e1e3d8911b25c2bd9ced973f9a
47002 F20101129_AAAFVC fernandezdiaz_j_Page_042.pro
8e3d70d3b0c1a6c7b67a942cf3c0c72c
a82e1ec0bf3f18c20d59985fe3dcf37fff3d2414
8692 F20101129_AAAGAJ fernandezdiaz_j_Page_098.QC.jpg
1a1cf62eac27d7508d494bd60b884bc8
3d6e9221ec8f6e573edf972854ecf14d852c3413
45398 F20101129_AAAFUP fernandezdiaz_j_Page_013.pro
e6b43aa8af12da3c2f0db6796d1a58b2
dda8bed2fd96d13559824dc99155b12aae7758db
30729 F20101129_AAAFVD fernandezdiaz_j_Page_044.pro
ae816d016a51715813259039c578a7ca
d5e941bfea8ddaa47c15d5323d7b4662145a94b0
6890 F20101129_AAAGAK fernandezdiaz_j_Page_057thm.jpg
0b70e5aca85b220a1a9980ea135b817d
e4a22453bfb5b28905c90352f27c0dd41d93d984
44827 F20101129_AAAFUQ fernandezdiaz_j_Page_014.pro
13b2f3163ac65061680416707153d66b
b51b8dd88442a319c75b61abfd2746641fdc5db2
31876 F20101129_AAAFVE fernandezdiaz_j_Page_045.pro
21c3e720b7ebc3f1a1915e4d81f6eeca
a5ffbf0b813506d1b86c8cff93d18d7d03f524ce
24567 F20101129_AAAGBA fernandezdiaz_j_Page_023.QC.jpg
ad5b4a46d852f74975fa0f69c1a503e8
a53f880b13f3717e81fbdecf31e65470a5094f6d
7373 F20101129_AAAGAL fernandezdiaz_j_Page_114thm.jpg
2961834dcad0fceee518387cf4f0f897
4ad85b3db8e90343ed50c554f2098ba1a782602d
36597 F20101129_AAAFVF fernandezdiaz_j_Page_046.pro
7d8487a081a0bf71269264e702405694
7551ba49cae5320daba9d6a6b981f9a04dc732f7
52757 F20101129_AAAFUR fernandezdiaz_j_Page_016.pro
638e1c97684999fff43efe5a98227f0c
41170ccae8753a73bbb81f7e32f4c152d9dd067c
25375 F20101129_AAAGBB fernandezdiaz_j_Page_084.QC.jpg
746e23dc38009b2e60191bc3a02286c6
111e2f8a44ef014476e46e78f0248db10763820b
6774 F20101129_AAAGAM fernandezdiaz_j_Page_019thm.jpg
bc9be3130f16da7167c782d78efe449d
c0d598bac72656b691bb81b07e8348b66519cc95
13448 F20101129_AAAFVG fernandezdiaz_j_Page_048.pro
50d7b2031ff4adcfefae6153850d39fe
421bb00b5db6e12713972bafe7a74753f8ab8e13
41665 F20101129_AAAFUS fernandezdiaz_j_Page_020.pro
657d14ac2ba598acfcb1c041e3465cb0
4b86bc960cacc84650f79dcd03f235a245b1fa65
23826 F20101129_AAAGBC fernandezdiaz_j_Page_075.QC.jpg
7ad7e5e25377dcf9fbbf3315d908ec8b
cf32400e4d9df241c47b0f73b6977e87c639c778
1356 F20101129_AAAGAN fernandezdiaz_j_Page_002thm.jpg
1491803bd07d85aa3a3e2c9b46011aa8
6663c3cf718b6d782031a6c659784fb81087a29b
27884 F20101129_AAAFVH fernandezdiaz_j_Page_049.pro
79f79857fb8a3744bb6dd23bba2479a8
a22284f51a66f0e7b75cdf81cbdb860ba0deaa29
51748 F20101129_AAAFUT fernandezdiaz_j_Page_021.pro
e8a39794ea98224c21487f57690486a9
073fc9b44d9c178dcfd8f72c4e8a7c8745b76814
6029 F20101129_AAAGBD fernandezdiaz_j_Page_090thm.jpg
c5f85d6a3870a765dca5de7dcb53f5a8
dbf136cc634c07a2869002d7ac0552abe80c1255
4898 F20101129_AAAGAO fernandezdiaz_j_Page_123thm.jpg
a766411450f98057bee477141ef28742
7d008f424d4ef40f42e6ad1d73ea402d867f4219
18693 F20101129_AAAFVI fernandezdiaz_j_Page_051.pro
69eb9c3da97c59e3873d802a745240a4
668af0448372f5a4b34cdd556b4fa98e5b41bcd3
55428 F20101129_AAAFUU fernandezdiaz_j_Page_022.pro
10c270aeb047ca4e5e33bdda72d4e31c
b46f269287d586578e5111dacbb4e54ae63c8485
5754 F20101129_AAAGBE fernandezdiaz_j_Page_113thm.jpg
62ce7da70fe9b9ce7c760f0cfbbecfaf
f4b70977fae683346d0843445ff875db29fc8d0f
7251 F20101129_AAAGAP fernandezdiaz_j_Page_001.QC.jpg
06c89d60425f8414128aec0dd0f3abcd
92029af931ab6b69ed50e0105e5c7b467e66931f
29041 F20101129_AAAFVJ fernandezdiaz_j_Page_052.pro
39e026ca389327d26def4d82e59171cf
099bd86e5c94066415dcae3558511fd98720131f
52097 F20101129_AAAFUV fernandezdiaz_j_Page_023.pro
c11ec57bcfae8845723e11f233da2048
18c857ccae67c5309e76664c40c918e49e82d1a9
23815 F20101129_AAAGBF fernandezdiaz_j_Page_080.QC.jpg
ba02003f324718f0092d0fde07fbd221
40caeb70330cf63c4ffae93b4e262ab1ae1c2145
6846 F20101129_AAAGAQ fernandezdiaz_j_Page_047thm.jpg
21e9cafd7353f55b5d836caea0abfe12
a8c7f02dd3978f753b07f974a41c60b0538f0b35
54751 F20101129_AAAFVK fernandezdiaz_j_Page_053.pro
b9aeccc217dec9b145a2fd6eac2092b2
e79c2f0fdae15da662b62cbc9ca9c6ea940b0006
43370 F20101129_AAAFUW fernandezdiaz_j_Page_027.pro
b2d64335061c7a37a671aca804bb444e
b2ff724e406a466aecb668be66d818bd3f0f3594
6358 F20101129_AAAGAR fernandezdiaz_j_Page_111thm.jpg
e07dff02395d1952c1d09f385431c2e1
758b9287a6a5d2eecb04210eab6d8f26d702e053
22217 F20101129_AAAFVL fernandezdiaz_j_Page_055.pro
57ac768fe0fda504cb290c73dfead5d1
aa28ce5344a35d5a9fce10908a7b718bb995d4a1
49894 F20101129_AAAFUX fernandezdiaz_j_Page_029.pro
c5e718d29508268b56e88b191b02519f
7ed8d8972f625461e2cc01a2024ed4ae397680d5
6600 F20101129_AAAGBG fernandezdiaz_j_Page_058thm.jpg
1b3db18a3d052fd96b5d0d234c95c44b
2dd4d0e2a1865947bf4f360ff875c9b91f0e8a4b
6838 F20101129_AAAGAS fernandezdiaz_j_Page_087thm.jpg
80fe0ba27dfe51037534cd09240f80b4
2811d128fd18ddf04e746a797b77b1803b874b99
38866 F20101129_AAAFWA fernandezdiaz_j_Page_083.pro
33c95a2465fd191fb64dd370a3e4c7bf
56c61accb4f8138a78014178e70ef086709aabdd
44799 F20101129_AAAFVM fernandezdiaz_j_Page_056.pro
d173cab80590dc99ef64a9809d6f552a
8cbdb586922d2e979b45c96419c18f664925f6ff
32617 F20101129_AAAFUY fernandezdiaz_j_Page_034.pro
9bff0862c2664f95e03af60c651dde97
efba28e813669bfa6432673bc09cfa40e8a30b2e
3733 F20101129_AAAGBH fernandezdiaz_j_Page_013thm.jpg
7c64b70929e42eba0e651b19c658d926
f86bfd98374dc03bb95e0ec0b388385cd37a6b47
3237 F20101129_AAAGAT fernandezdiaz_j_Page_133thm.jpg
67a4a58bb938c49ffef6d1bf02004897
e87aee12d170e71a78b558a80e8c3e38ea715484
52372 F20101129_AAAFWB fernandezdiaz_j_Page_085.pro
e99c5deb37c62d04a1e2fb4d382b2ad3
a2bb6f2f3dca9a16c1a8d2217c40ac01e5505116
22362 F20101129_AAAFVN fernandezdiaz_j_Page_059.pro
ed9af868ec9d13b95b2e034010858e6c
58de6b4b1dc7691a23dc7c6d28b77ca571914eb2
63985 F20101129_AAAFUZ fernandezdiaz_j_Page_036.pro
03f8ef38874166402a2dbeeead66d172
d7165720dc54e830df9108d16c6dc303ff435978
26005 F20101129_AAAGBI fernandezdiaz_j_Page_100.QC.jpg
51eb23ae99f62ec5b6831c51f8ab562c
e2a1de0195dd87c017f51ad5c1a1406a7a685606
6127 F20101129_AAAGAU fernandezdiaz_j_Page_092thm.jpg
0592ce008f994a15ff3b7fb4706850e2
fdd59d9fc60ce6b2593a4aefba3889a6dced0842
52224 F20101129_AAAFWC fernandezdiaz_j_Page_086.pro
b911b1a9f5d2436ad98925c4c08d9533
82a397e4d3b0174af0258a9b3bc980b0bfcaea11
31942 F20101129_AAAFVO fernandezdiaz_j_Page_060.pro
73d7af9598378db6b78e38c57960c73f
5caacc83c850cb7a29f140bec2e4ae182614175f
6813 F20101129_AAAGBJ fernandezdiaz_j_Page_104thm.jpg
c984c95d766420e98c8694a891a88f32
848f985e61507fa8a3aabf1ad200d8668493a25f
2413 F20101129_AAAGAV fernandezdiaz_j_Page_115thm.jpg
7d100e5b2fbb32f124d83a8644444933
b570d8f29010fe75c156dc0ae63ce624229e1b86
50115 F20101129_AAAFWD fernandezdiaz_j_Page_091.pro
4e9ed09cfaca3d81a798e2db17a151c5
8c86f4bd0e6e1057cae5578f881ac051566a6661
49134 F20101129_AAAFVP fernandezdiaz_j_Page_061.pro
ba7a9617901b0c04c716085bbed7af52
8913df016d9cfb190b0d6842764931931107cacb
13995 F20101129_AAAGBK fernandezdiaz_j_Page_004.QC.jpg
36a5913ad3dff0c37327c659c73d7216
55ee9f31eb9cd455e897f0940772532942a89c3f
25227 F20101129_AAAGAW fernandezdiaz_j_Page_057.QC.jpg
99e262b9cf9de55e7fc953ad131f2951
854308eda11fd1a19a745a5820571c852b56960c
31371 F20101129_AAAFWE fernandezdiaz_j_Page_092.pro
cbc19e6cbbe61f07376d9702ea6277a5
617f887077cf3daef0a067dc869381446c1145a4
4724 F20101129_AAAFVQ fernandezdiaz_j_Page_064.pro
b707c3198c9ccb254a41cfbfa79e8374
85d6245a506258c8ce3f5e052dc3817210858305
23642 F20101129_AAAGCA fernandezdiaz_j_Page_095.QC.jpg
d87ad405d91dd0586ac84b681ccac786
24e16f56ea2ea0377c63e131d9d90f9423c6deeb
21922 F20101129_AAAGBL fernandezdiaz_j_Page_096.QC.jpg
2588e3a2d15771d0e38c53964edf3e80
b669212b2754b100cb6479027a6909aa5012634e
7218 F20101129_AAAGAX fernandezdiaz_j_Page_054thm.jpg
b0068aaff422c644ae8141f6af173d02
5860aeee831b83e03e1d92bb590ae0b0268821d7
19609 F20101129_AAAFWF fernandezdiaz_j_Page_093.pro
8c05fef9660378c28396fb932d4c5565
4e6087cfd077708df4769e91077b04ce5ec7f2f5
44022 F20101129_AAAFVR fernandezdiaz_j_Page_066.pro
78d23b950069549804149444d1ff50dd
be47d6fff0d96069c9f725dc119e11c7f7b34c60
17415 F20101129_AAAGCB fernandezdiaz_j_Page_033.QC.jpg
c0aa58e413fccdd94fb34c2adfe4b1e3
647ea0fc9dbc68cb89ef57a31374ddaacbb1baa2
28777 F20101129_AAAGBM fernandezdiaz_j_Page_066.QC.jpg
63a1a0967255eb1c3a4172cad98bc055
106687a01d203f0bdd6b8aedb4f488bf9b60a118
25083 F20101129_AAAGAY fernandezdiaz_j_Page_067.QC.jpg
b6b49aca127c2942786f5c36f824d348
41903a50bf05edbb2ce3ca038ec18627a7733d02
35983 F20101129_AAAFWG fernandezdiaz_j_Page_094.pro
75141ddde236a7c55b925e0712627156
acefb5fdc9a7d318942e45fe2e11dc11ed56aa83
52121 F20101129_AAAFVS fernandezdiaz_j_Page_067.pro
3376c4f0bd1f9f048f5ea9677962f77b
979221535d170657ee92372e13886d97644bede6
7693 F20101129_AAAGCC fernandezdiaz_j_Page_129thm.jpg
15e1273b6624cea96fd11a2da0738175
72410d6211a6caea2f8089d6523fcb80f9a4ce49
7139 F20101129_AAAGBN fernandezdiaz_j_Page_100thm.jpg
a24e43cfff18c2353107a7e22cf50851
afde30632eb49330ac0a0d942e772b17c3e2bda4
23142 F20101129_AAAGAZ fernandezdiaz_j_Page_118.QC.jpg
a66c1725ae918dc17204f9422226d21a
8f5a6a8b70231a53f44c75229716a24ebee9eb77
57461 F20101129_AAAFWH fernandezdiaz_j_Page_095.pro
943f7cfb26f5ef8baad413ff4f94ce8c
4efcf795ab351163b09e087c7e34aa0ab31e79ee
12827 F20101129_AAAFVT fernandezdiaz_j_Page_068.pro
7176b36f415ee7b512f579f9ac89a83e
8f46fd5b3a30d522ec0073f6de374edbdcd9e639
26160 F20101129_AAAGCD fernandezdiaz_j_Page_101.QC.jpg
ada09641d666ded7b14e57d8093054fc
51d90c3872983e58a2d5537d3828d73176cc86e0
22678 F20101129_AAAGBO fernandezdiaz_j_Page_058.QC.jpg
1cc366be3fe11c749032a0ba7d829c9e
c522564c04f927a8d73e268d2b552f6666b997a5
32650 F20101129_AAAFWI fernandezdiaz_j_Page_096.pro
00ba9f1c44c622a728db2385592304c3
72e3ff5977e198eabd4256eb95f4bb256978c0eb
42629 F20101129_AAAFVU fernandezdiaz_j_Page_069.pro
5441b6c159f2a7b082b946cc6fe16405
23ef299224ea31ff5d24f2532f7c8c100a42e149
6423 F20101129_AAAGCE fernandezdiaz_j_Page_037thm.jpg
0b48166819cd92cf7dc5a7ee3d03c073
f695448430de19def89cbebc1d80bbe3edce2c4a
7536 F20101129_AAAGBP fernandezdiaz_j_Page_078thm.jpg
eefcc7ee667d099f600d57d422fc1fc1
178a0a5eed2ea610f79158fd72a1add99694dd54
66492 F20101129_AAAFWJ fernandezdiaz_j_Page_097.pro
e6c5cf75409b65113474d15ea319ae68
9d683abf7f379d4187adfe81a55a39258ad9eaeb
41544 F20101129_AAAFVV fernandezdiaz_j_Page_071.pro
38f47bf66e4e6a4890bc7bf9e4897779
4cd1474aa76e7da7f2622dc5fc53626a11d70dcf
205848 F20101129_AAAGCF UFE0021101_00001.xml
f548c8eedc61365f285e9f6a2a4b4e38
5089f4cb6b039a40f75e2a0d2eeaa58322ff9d8e
5971 F20101129_AAAGBQ fernandezdiaz_j_Page_091thm.jpg
da52875c1faa7d0b31374977322577ce
f0e33dd3e468914598bd1b4fe7ceab421c33c1a0
56273 F20101129_AAAFWK fernandezdiaz_j_Page_100.pro
ca4fdf7f7a860244f8590ff646d413cb
9003d49ffc24e547f1c845f06a4d9bebd58353a1
50698 F20101129_AAAFVW fernandezdiaz_j_Page_074.pro
684e53e0bf59f98b500e5f31e6ac4016
75dca8a5b15acc9c0ea0170cc491b1c2d028919b
22577 F20101129_AAAGCG fernandezdiaz_j_Page_005.QC.jpg
e768669f1637a752e42c8ba0a7736a09
a8213a24e4ba4fd7b1650dbc63c0c9d760b8dac6
7479 F20101129_AAAGBR fernandezdiaz_j_Page_064thm.jpg
3cf9bc894bc51d97bc3bef46e720733b
3bf258b07f30bcb66d8673e6305b59a85ebaa7d6
49979 F20101129_AAAFWL fernandezdiaz_j_Page_102.pro
16d07238635234e7fc1a17b66f0b225d
035be3a02bea80dcb82c952dfef7c0a30244095b
48517 F20101129_AAAFVX fernandezdiaz_j_Page_075.pro
56e089143a59d751f32ba75924f95ead
3d32fbf573b8eaca767e2efc700b6323b6713cef
17362 F20101129_AAAGBS fernandezdiaz_j_Page_034.QC.jpg
84ae985cd59f9c7768bb791bcf517adf
f6ca4b33b0ac0a95d1308c49384524d99070d743
56481 F20101129_AAAFWM fernandezdiaz_j_Page_105.pro
cf4caace3d0b1f7db8e186c338b35fe2
a53e431d3cd4c505016f49f57f0492210875d748
38649 F20101129_AAAFVY fernandezdiaz_j_Page_077.pro
030db5cb0c9825bc86ec49a78012a860
94d73245ec540cf37b7a3bb6faaeb1f26bbe7713
64642 F20101129_AAAFXA fernandezdiaz_j_Page_128.pro
092cf832d17a5bd17d3630ef03bfc477
8b2c28064420a6b013ddb8e0be00d1c19198b2ca
5518 F20101129_AAAGCH fernandezdiaz_j_Page_005thm.jpg
64019fdab00406b0bdeb3dc1652babdd
6187cbdc38055b1c4485ddd1a5799f2be7beb7b1
7273 F20101129_AAAGBT fernandezdiaz_j_Page_032thm.jpg
01a4d280d3d5d912d21a9019aa0c6895
3eaa682eed5fb13370590000654fe6ff30428cae
6803 F20101129_AAAFWN fernandezdiaz_j_Page_106.pro
30b417b5c23342c5946736a8be80cdd0
92a474a6ac9530a2cf003f6b4b6a29ebe7ed20fa
51800 F20101129_AAAFVZ fernandezdiaz_j_Page_080.pro
aed3218d5155c058c91a455fd07e2538
a9d302148700da797180d23beafd6eb9195bdba6
70045 F20101129_AAAFXB fernandezdiaz_j_Page_130.pro
494d176ee70cf152d7281def08ae086d
1bb12f7fb2c60756045fa6d052c4b643ec9a0bc0
23385 F20101129_AAAGCI fernandezdiaz_j_Page_006.QC.jpg
feac54e2292cbd0863ddf5e0243aca76
75a8202aebc9e7a80bc3c7f17c7d372c20ac9bc4
5782 F20101129_AAAGBU fernandezdiaz_j_Page_006thm.jpg
e10c22e46b0eb7d7f098c43f5595462c
296779cec1d70f173dbdb6f017d3dc7bf6fa9917
10391 F20101129_AAAFWO fernandezdiaz_j_Page_107.pro
0b7ae4c226149b615f1bf46abd1f2c34
f0ec065e7345f47aa04246eccf4580224f9a17f8
13384 F20101129_AAAFXC fernandezdiaz_j_Page_131.pro
18594d7d7b882d659113be0d1d8ecf57
25a433b5a1634b67225f06ee9ab5e683d20b03a8
25093 F20101129_AAAGCJ fernandezdiaz_j_Page_007.QC.jpg
c1d910ab2ce92c1343c6dbe6230994a9
815b094a6f371ac672769cd7de382f9ff93b8a03
5926 F20101129_AAAGBV fernandezdiaz_j_Page_082thm.jpg
3b6541c1c5b12f1d979edb7307c7b3b2
0b9e4ccce435b2dc44784e6f7c3c4dcc46f1df69
5163 F20101129_AAAFWP fernandezdiaz_j_Page_109.pro
feabcab27f175df7b6eb06aca8eb9bb9
e0648f4c3c90f4fa2e3e49536460994e65affd8c
18874 F20101129_AAAFXD fernandezdiaz_j_Page_133.pro
d2913a0281dd0893f50c1fa9e7768904
56aab492dc001d2aecc0dcaff317b93b2f3ed5fd
6339 F20101129_AAAGCK fernandezdiaz_j_Page_007thm.jpg
b731095d5edb3c37b40dc32bcfec206a
6dd77a9b26328f2add47f73e5e5b1404c974e3f2
15344 F20101129_AAAGBW fernandezdiaz_j_Page_123.QC.jpg
4292b967bae3315686c8c060ae5a307d
0c54cfd6e80a0aba908a8ee0f2ecaf86aeddfd3e
48690 F20101129_AAAFWQ fernandezdiaz_j_Page_113.pro
59cf7a25e88d2edb16683eea9f6f8478
6dc091b3a546ef3a4e1bd77a64f8a5b6d58b2b44
490 F20101129_AAAFXE fernandezdiaz_j_Page_001.txt
3051efa144a587628f78ab4d2a7e51f7
9c8a3798385c7351ff23457fc72850233feecc96
6750 F20101129_AAAGDA fernandezdiaz_j_Page_023thm.jpg
fb51fc2f242b71232bd265f1d7f7e2d9
1e6e2ee2544de48bec28af38b85c0cf08f6cfa7d
6304 F20101129_AAAGCL fernandezdiaz_j_Page_008.QC.jpg
430163e7815dc609c855fc9c3c2c4d2d
72a1aa83fce257d94c3ed75d06356140c46a5fcd
6108 F20101129_AAAGBX fernandezdiaz_j_Page_056thm.jpg
053068b2e03268696b66c94c41bdac9a
601e80be588510f17e4022e31292bc30a3676e70
41606 F20101129_AAAFWR fernandezdiaz_j_Page_114.pro
033e7c4da5aed2326950eec4fabb6ca6
48eb4a944173d7a15871652faa1965ef3d042daa
100 F20101129_AAAFXF fernandezdiaz_j_Page_002.txt
98384ff5c961376b0cd5602729bc80a9
81907f94d96435e15a062964dc85acce1c6054fa
6850 F20101129_AAAGDB fernandezdiaz_j_Page_026thm.jpg
9b2e64e6d677389a9672a83b2013f0e8
d00a59f6eb4abfba7e34a766b6b70977e86fead5
2051 F20101129_AAAGCM fernandezdiaz_j_Page_008thm.jpg
f1130a004e8608a2ccbefe351337bbc7
fe6597ad4a90ecd591dce37b40bf02e1b753c653
7306 F20101129_AAAGBY fernandezdiaz_j_Page_052thm.jpg
ba09fd9cc7bc7d42ef8c5abe4c812d18
4cc4630ab8bdbd7de572849ab13a8e0db1baeb38
51743 F20101129_AAAFWS fernandezdiaz_j_Page_116.pro
dce6abe9d1a09f76ac16986f4dbd09c8
6bdc40709303e223a007925b2ab99836e49f37e3
286 F20101129_AAAFXG fernandezdiaz_j_Page_003.txt
46f3fd448ce0f565eded5ad5f3273be1
a310eea3276fa9983b3bd14bca1a125053080b15
26949 F20101129_AAAGDC fernandezdiaz_j_Page_027.QC.jpg
6c1fe194f2053a189f200b4f7b7afa8a
6f04af4d9807c00ddc263bad050b29f778714594
5911 F20101129_AAAGCN fernandezdiaz_j_Page_010thm.jpg
019ccad3a44ca348c1aa9128647281ac
6cef5f47f50be09c073339c8a432590d8a04d071
24950 F20101129_AAAGBZ fernandezdiaz_j_Page_028.QC.jpg
9bda8841b7048661625ce2f52f760999
d040776457b886d08534dacdb067fbcb45385940
52830 F20101129_AAAFWT fernandezdiaz_j_Page_117.pro
01a7145d03c6cfa91a4f280aa359a798
4bfe011b2e195b8136034f1c85ce4d2c912dffd0
52026 F20101129_AAAFAA fernandezdiaz_j_Page_082.pro
43bd4c2a2ca907dd681b80638cdc61a9
5ac200990ddfd89c28d261bb05499ae053e69cae
3894 F20101129_AAAFXH fernandezdiaz_j_Page_005.txt
bd560540a0c6e21d09c811a4ff66a300
4a9238c0dff942efb49dd45aa26977d304355d1f
6726 F20101129_AAAGDD fernandezdiaz_j_Page_028thm.jpg
194944ea5ed2a1bdfecfd6285ff7c4a4
e7d76948318453f5d5d670b03772fce0c1fc7063
25483 F20101129_AAAGCO fernandezdiaz_j_Page_011.QC.jpg
d1191584b18a540eb8f8b4f8545677ef
c9b4f28ffe1f2c43afe2db8c66228282d947fcfa
48864 F20101129_AAAFWU fernandezdiaz_j_Page_118.pro
0c08639de0834c4c36c59d3d0a5d46b9
36232dfc51948cbc5014ad3eb408f325cdd72622
68225 F20101129_AAAFAB fernandezdiaz_j_Page_129.pro
f5c5ee33372962cd668cd5650482ad19
88e0b8f9410cb2c7eae0dc29ad769915a08c530b
758 F20101129_AAAFXI fernandezdiaz_j_Page_008.txt
4545886eb7f3d8426bf70d3fa988f728
6f512fac4c4ba129c286a99fb90c15705f202d65
6669 F20101129_AAAGDE fernandezdiaz_j_Page_029thm.jpg
eae9d982faeae36d480158f9aed96a7e
7bef1b1e7c5f68de2739491df8123e1fd29e3fc8
6406 F20101129_AAAGCP fernandezdiaz_j_Page_011thm.jpg
d4c6b8451c4ed2a776bce71166bdcbfd
7c98f3671cdc88eaf5f93d913f24dc4bdeed4286
25025 F20101129_AAAFWV fernandezdiaz_j_Page_119.pro
5eea72eff77d17ef0a69ef2526c8abe0
77f129118c8e02cfefa235ae328917588c1b1d0d
78490 F20101129_AAAFAC fernandezdiaz_j_Page_026.jpg
0518904e50f9d86d12b953b9a8da7c20
5b7a3e7bbf3bd66d3c405551761543d0690f7ecb
2601 F20101129_AAAFXJ fernandezdiaz_j_Page_009.txt
f42275ad75f2bca05603fe22630cc791
8e0dcf7357533b753c596e2648845ad19b05e6ea
22469 F20101129_AAAGDF fernandezdiaz_j_Page_031.QC.jpg
9cafe2290185138ba15792ce4c44a53c
54bf5cecd38666dce0e9c2aaccd46306b7eb0caf
6851 F20101129_AAAGCQ fernandezdiaz_j_Page_012thm.jpg
d68679988816db3db3d50e50471e2a79
44ecc42ad2003d331162b670dc93a326eb051421
21644 F20101129_AAAFWW fernandezdiaz_j_Page_122.pro
eedcfe7a9b21049bc16dc0a39035a338
0a848222ef13841b37fd9901d10d7bd8800fa595
1765 F20101129_AAAFAD fernandezdiaz_j_Page_072.txt
94baed385ba64a01dd6cb00055711827
2b7d73c52ba4ba956692866168b1a3fc23f10a47
2960 F20101129_AAAFXK fernandezdiaz_j_Page_010.txt
c7d95526299a1d32823cf29b4d50bba3
2cce85e0b99e04ab965db639a76ca035310eea75
25789 F20101129_AAAGDG fernandezdiaz_j_Page_032.QC.jpg
a1c16d870fbc3a210561602d6c9d3f96
548d62524b5ed7215866ccb5ce0fc7b7953cc1b3
5974 F20101129_AAAGCR fernandezdiaz_j_Page_014thm.jpg
d3149004b405c404840c43e1876d680b
569547705965976bc7afa086d6527cfa28146ad4
21907 F20101129_AAAFWX fernandezdiaz_j_Page_123.pro
2a589632faba77f02ecfab65c944ff50
7d8426a3cffe7c8d0632904180d4eee16192b373
F20101129_AAAFAE fernandezdiaz_j_Page_056.tif
9cd8fd5dac4aabdea344ca051b7724a1
4f0ae7ed270c019935c75919922e99f03129fad8
2923 F20101129_AAAFXL fernandezdiaz_j_Page_011.txt
86ce5a3739dc3c62869fd2dd19324a74
f1c5619dd183abddbf457f81e928b69605e3692a
5383 F20101129_AAAGDH fernandezdiaz_j_Page_033thm.jpg
35a3c034a4853e3139718182413f26a0
8c7d95e451366f70d924aca51cdad0e9d36a357d
15743 F20101129_AAAGCS fernandezdiaz_j_Page_015.QC.jpg
9e5bfb098b09916d41707e2dc69033cb
b59d6d57a6bcf3dcedacdcc76cd0a461ac969647
70110 F20101129_AAAFWY fernandezdiaz_j_Page_126.pro
4d2599b13bb909586986195b46e59b56
5f13d042994bdecba847c239fd98cd7342e8c51b
2036 F20101129_AAAFAF fernandezdiaz_j_Page_061.txt
f54abb6c6e408fe4505ae6c0abbf55ff
143739c437142e6acc858e37590a4499ba8061cd
1864 F20101129_AAAFYA fernandezdiaz_j_Page_037.txt
b410dcc22502e009e1c37d11cbb0b861
b7b424b36efeb95b9cab5bc90991b0c6008dc7f7
2909 F20101129_AAAFXM fernandezdiaz_j_Page_012.txt
8d5e131408af509a1800b4d5ccb09ef4
166296897523fdc3c4024fe67f35c98eed006ca7
4408 F20101129_AAAGCT fernandezdiaz_j_Page_015thm.jpg
396fc9a3766a1f970ccd6cfb51730471
10da5157aea975e256b4ed7420ade944eb49930a
70604 F20101129_AAAFWZ fernandezdiaz_j_Page_127.pro
9b1d141f1627416b2b53462128c85ecc
889a4a975601cb476028c2733b0adafdf6fbec0f
1537 F20101129_AAAFYB fernandezdiaz_j_Page_038.txt
8547d1d9075975cb50363b6c13530ad2
ac8024bd6544698fe10cdb1a87fd2ea220d9ffd9
1973 F20101129_AAAFXN fernandezdiaz_j_Page_014.txt
2cc1e4990a2148d39b67633afd60b017
9b55208a402d6b4bd356300de24b05f89e765f29
5171 F20101129_AAAGDI fernandezdiaz_j_Page_034thm.jpg
200c6272267ccf62b7724b874307d624
d76f39324e7b143f09dc970d72f84579a9e6f462
24974 F20101129_AAAGCU fernandezdiaz_j_Page_016.QC.jpg
e11886a2474caa8f1e554f305b7dfa89
1448ab7a02dfdc33592e59aefe95af38bee6ab43
23349 F20101129_AAAFAG fernandezdiaz_j_Page_042.QC.jpg
6398769c8dafcde7a2e55548d7f4df46
06bce11d984c03a41a030baefb3abec86e4e644a
2142 F20101129_AAAFYC fernandezdiaz_j_Page_040.txt
9a6b79a1cc18fbea7231ac50c6144f66
d225342fbacac1246a4a54ccdf21be17b8df4a7b
2149 F20101129_AAAFXO fernandezdiaz_j_Page_016.txt
acc4aec5d58a0758bc9777bb6919442f
a9146d383723be72bcd7d4cbb9827c001f701d89
6547 F20101129_AAAGDJ fernandezdiaz_j_Page_036thm.jpg
2dac958491a0832b6348349fcfec551e
93a91c937fa8ddf42f9f68c4864adf23cafb58df
7016 F20101129_AAAGCV fernandezdiaz_j_Page_016thm.jpg
eb5d9bb7cd92b8c0e27abbbeeea44c84
6cc9c6092417ac5c6b1abb552172931a1eb6a816
2067 F20101129_AAAFAH fernandezdiaz_j_Page_078.txt
78dcdb537dd488e7bd57c992b4736a0b
5c1999989591d53bddcf1baead1d2de035d23384
2205 F20101129_AAAFYD fernandezdiaz_j_Page_042.txt
9a135d04b8a3093cf91bbc5c2384575a
8e5c123b15e8969df4d00feffd0e53bad0b5a878
1845 F20101129_AAAFXP fernandezdiaz_j_Page_017.txt
46064e967821910c224a5563df401d12
677d0c6a0d03d77e9a2b2a0082269b727366ed9d
22123 F20101129_AAAGDK fernandezdiaz_j_Page_037.QC.jpg
2a6e1228a891e1a317653713ecac4b3f
3053e362f635860c30e1f25c6559791216012d3b
25986 F20101129_AAAGCW fernandezdiaz_j_Page_020.QC.jpg
a6173e953880623c916efbbc512af9da
5599016598ec9413b860eb4caa8fa0eef4f489fd
36782 F20101129_AAAFAI fernandezdiaz_j_Page_041.pro
0e731ca673fcb4d84b4958944b395e9d
9b53f44f8b92647be3113c26e205e5e4568aa8bd
2254 F20101129_AAAFYE fernandezdiaz_j_Page_043.txt
882e0b14b93437d210f58fb0bdfba458
def0ff4690ebe3a9c0d44ae484563930c4151f8f
1685 F20101129_AAAFXQ fernandezdiaz_j_Page_020.txt
2824ff6fa9ad52a85c81af057ee5c2ec
4ea0c3416f97cffb5b625f5a07de25eb0129edb9
21700 F20101129_AAAGEA fernandezdiaz_j_Page_056.QC.jpg
e8b4c59b7c9be417ca175ff7da1bf8fd
8ebd1a81ee26052dc12ca26e4bb58094ff6d87a6
25908 F20101129_AAAGDL fernandezdiaz_j_Page_040.QC.jpg
dd848eba9d1eb64ab5238f928c860e2b
0505c5a8e423c46ba56ea20e52ba126c489e6af5
25117 F20101129_AAAGCX fernandezdiaz_j_Page_021.QC.jpg
08fdd48cc45b7aca73e58c3bafb67a12
addfb9c32ac091a195d263e18f7c5d71043f2f6b
22257 F20101129_AAAFAJ fernandezdiaz_j_Page_094.QC.jpg
588c8f819e250b264a881574e16fea54
28a46ecbe18d565f946f853a4bbd5974ad995e64
1218 F20101129_AAAFYF fernandezdiaz_j_Page_044.txt
a45eb14cdc7c56c631c0a83d642108c2
e191fb613c51845a214e6873219b461b011c60b0
2049 F20101129_AAAFXR fernandezdiaz_j_Page_021.txt
2190d9a530166a88769a0ef4ea75d8f3
18e76f5ceaa1fab5e8b931e9f919f62ac3db5620
28103 F20101129_AAAGEB fernandezdiaz_j_Page_059.QC.jpg
809c1668655874127d294d2800a18102
57ae1acddc83f148717322ea2835965492a6fd77
21929 F20101129_AAAGDM fernandezdiaz_j_Page_041.QC.jpg
ffcf8404b1dfd318c119edb228d166b7
f8f76e488073f4963c04715b849e2d55db26efd9
6951 F20101129_AAAGCY fernandezdiaz_j_Page_021thm.jpg
5dba3a5bdbc8a2466a98b94c68734cef
15f80a31f22fc86a6d1bd12f2788b8b847f0c2c1
114005 F20101129_AAAFAK fernandezdiaz_j_Page_062.jp2
b75fba11ed9aaa136d7c4ca7d2d674f1
7b68d43d673dd25852f9495c2af2aa3c2e9b109c
1489 F20101129_AAAFYG fernandezdiaz_j_Page_046.txt
8b9fc19929cf89a210fd2dbc6f1ea3ef
ad024d8fa0efbecdb3dd4a47b2ae1fba41f61e67
2174 F20101129_AAAFXS fernandezdiaz_j_Page_022.txt
de61bdf2f2ee49cd028bd7a673c006d2
161ac0ed20736a15a48411562f92c7b7d782e69a
22131 F20101129_AAAGEC fernandezdiaz_j_Page_060.QC.jpg
2784710070f751ca15a4b3e98e5eb1d1
3e0df974f64568dba60c4a299b91d6cd97a8c49e
6453 F20101129_AAAGDN fernandezdiaz_j_Page_041thm.jpg
38484d9534ab892014d791e9bbea7ca0
1a2f1197a71552a9043bff91639d2416096e98e9
7138 F20101129_AAAGCZ fernandezdiaz_j_Page_022thm.jpg
76fd7b5fe6016b634c9e3065daeb2ecb
6ab4e69f8efeb61b91d95ea9789ea65ac11e0ea5
7030 F20101129_AAAFAL fernandezdiaz_j_Page_068thm.jpg
d031197de138f4d0df4a4396173260e5
7941dd90c43780b74630aa7caf2a03789ed7954e
2005 F20101129_AAAFYH fernandezdiaz_j_Page_047.txt
4e75a91af006577f8e944e3c2dfedfa4
d130fbff40ee3b6301dcd4031438100798be9554
2057 F20101129_AAAFXT fernandezdiaz_j_Page_023.txt
80406992322a904fd8cee262a82cb155
a4ee9a0367d17db1fe5e05ce945101f642320555
53768 F20101129_AAAFBA fernandezdiaz_j_Page_018.pro
bd51f5fbbfceac84866cbe6406a16364
d4b75925433ee484f42d8d1b685d93e6db89b726
6682 F20101129_AAAGED fernandezdiaz_j_Page_060thm.jpg
18b4bec02862401f3ca9265722a15cba
0f6f1ea8e4188370e46a247dae6f6bf0f892e42e
6565 F20101129_AAAGDO fernandezdiaz_j_Page_042thm.jpg
3134240053bbf0c921433ab73cf9a280
ce940a58eb1c13419efca6f1b14c7b81dd9dbc54
33407 F20101129_AAAFAM fernandezdiaz_j_Page_031.pro
beab3c96642a657a735a780a9dbaead5
82dd84a0d464579a910d23559d037558951e8009
1142 F20101129_AAAFYI fernandezdiaz_j_Page_049.txt
019a1ddf4d32fb6d7f0b6e147f99063c
4f07f4c3271a6079bd4fe50e6c4571bd5cdb548a
2024 F20101129_AAAFXU fernandezdiaz_j_Page_026.txt
c80e0d9970fde0c69575f3158461a5dd
cd106bcd6cb7b76795eec25b5f38f4fc32eb30a9
26020 F20101129_AAAFBB fernandezdiaz_j_Page_026.QC.jpg
463a08bb3406ae3dd77728b0b92509c0
095e77a92780bd2eeb386eeb5cf3e9936494f502
23817 F20101129_AAAGEE fernandezdiaz_j_Page_061.QC.jpg
53ba6d7186b786ce5e6d0841c9a21f1b
ad1c320c529938f26e9b35581e5f4455a5041eb3
7215 F20101129_AAAGDP fernandezdiaz_j_Page_043thm.jpg
3af906d1e47208c2aedc48e437a039f0
86a7a0ae14a5a31ba9095238e2044add8c93a988
7013 F20101129_AAAFAN fernandezdiaz_j_Page_097thm.jpg
e89dd2eec81876b08c342d33bfad73d5
7c73aa4e7dbd0025dae1e2d7864c63b9f2d3dede
2155 F20101129_AAAFYJ fernandezdiaz_j_Page_050.txt
06772c1bafe12539ab4a9aac530c372f
c1d98e992acfc8cb5099de9b841c88ec266c1586
1756 F20101129_AAAFXV fernandezdiaz_j_Page_027.txt
1d2507385fd17b1aeb9f5365874358d0
1c4c063bb382fffdc58a15d1234bc6395268b5aa
1051980 F20101129_AAAFBC fernandezdiaz_j_Page_012.jp2
4ffa236da22abceddebd97e6a9a1e9a4
4cdb074c9a03b9e8e11ecd4f3be914caf5173183
6430 F20101129_AAAGEF fernandezdiaz_j_Page_061thm.jpg
0750377dbbb1891f4a19e1568f1e8735
9ac8628a9b47aa4fb7503dd91395c5133f04ec9e
23923 F20101129_AAAGDQ fernandezdiaz_j_Page_044.QC.jpg
ff5657ecd3dde08c3e03e16b5228054e
d17d41a57a0bce0daa28e3d86730a82bdd8efbba
6658 F20101129_AAAFAO fernandezdiaz_j_Page_107thm.jpg
88e821c152418dea36b93e8ce8a00556
22b2a87bba070ca3bf9038e2688f819dfd1d1ba1
777 F20101129_AAAFYK fernandezdiaz_j_Page_051.txt
2d918b0afa430013efccd022716379c7
9f31334d2d62a1eefb254d401d85e46762187c89
2043 F20101129_AAAFXW fernandezdiaz_j_Page_028.txt
b66f68e21fab38297bd613c431acfbf9
4017c74a70f7ea103caaa04b022d5d1031894015
68118 F20101129_AAAFBD fernandezdiaz_j_Page_083.jpg
3ad613cc64febcbfdeba3905678890c7
85f3d4bfabc3a3715acd46a7103b12e41a19918b
26306 F20101129_AAAGEG fernandezdiaz_j_Page_063.QC.jpg
40f2fae5faa61e657790801b07212d55
7e337bb540922e4a8838f5be84ed8cd345ce8b07
24822 F20101129_AAAGDR fernandezdiaz_j_Page_045.QC.jpg
d95babb86efe691f88d7d7179d1465b4
2671b4cf246fa28c41d3139f81d9a534736e3095
F20101129_AAAFAP fernandezdiaz_j_Page_078.jp2
7687986148179a85c44bd6a5ad1c798e
4c737b6592fb999118d1cfb9ef57f9d49f33aaf1
1364 F20101129_AAAFYL fernandezdiaz_j_Page_054.txt
cea5142ec05e680042f985f429b43b3a
93d0a501c6260a8996a06b6ec070713690165036
1987 F20101129_AAAFXX fernandezdiaz_j_Page_029.txt
0a0b2b3014ac13836964f33d0455025e
1f3a075402beb0913bd4d3e48b42b207df45eb8c
2114 F20101129_AAAFBE fernandezdiaz_j_Page_079.txt
ca3ffbc9d2d6c4077c6bc8ecc1dacfe0
a84b7228ff8cf1de7f2d916ea0c3c498763f2808
7461 F20101129_AAAGEH fernandezdiaz_j_Page_063thm.jpg
16beb14f9077db79f98bb45db0ecb224
059ce0aec9b1cf61a625b6b3a196aad39944e8a6
17352 F20101129_AAAGDS fernandezdiaz_j_Page_048.QC.jpg
3bdb2ba4e949112afbe55c1ac9f0b606
c704dbacfff7098d5fa7f0085526c1be03107ebf
7534 F20101129_AAAFAQ fernandezdiaz_j_Page_024thm.jpg
4ede200d453f00bf4cd825e2dc35aa29
fa583b0d9c27476f68b2cd5a2914c32c998ccec2
1358 F20101129_AAAFZA fernandezdiaz_j_Page_081.txt
94dfee6a049dd7df798ec39bea82e31a
47f7fe3fb1543451f857f7fc5f70db027fd642c9
894 F20101129_AAAFYM fernandezdiaz_j_Page_055.txt
6da023ff9aa335ca1ec704d62ba210e6
58ef3ea47d70843b3dcaa50e29a4bca97737a977
1776 F20101129_AAAFXY fernandezdiaz_j_Page_032.txt
6e0b760c034d2750c9bcb2750528fafc
5a499743d6f9bbe0c59a31a0a91bb74d9e87584e
6621 F20101129_AAAFBF fernandezdiaz_j_Page_062thm.jpg
0d9058da4bb02b84d4a0fa0d58d10522
6fe46b4a8231f5437424fee5d1e1dd5caaf4efe4
24394 F20101129_AAAGEI fernandezdiaz_j_Page_064.QC.jpg
f22e7b29d69cde16a5cffaf70784f472
482298a9c7b5f301c9ac4e4b103aa75407d891fa
5664 F20101129_AAAGDT fernandezdiaz_j_Page_048thm.jpg
be10a8ba4d917c427ac6cd3d0b8b4412
0bb6ec9b9df547de0a04676169673f2dd65b68a9
63067 F20101129_AAAFAR fernandezdiaz_j_Page_113.jpg
41bf145b11d26d0a5d742a3ac271f047
78cce4f8f2534b468dd5bd9465d4dcd28de72a67
2096 F20101129_AAAFZB fernandezdiaz_j_Page_085.txt
b5302ec0b39f4112f1a95e5ea1be970a
529298efb2ab99ed46c863e9b039072f6eac93f8
1789 F20101129_AAAFYN fernandezdiaz_j_Page_056.txt
2321e4d738c8ea44626ee83bd627f487
183dbaefbf15522cde24983357431eb30a180943
1907 F20101129_AAAFXZ fernandezdiaz_j_Page_035.txt
12efa5a0ae9fdecad36fc84dda07d39c
2e25cfc2e68f5a7268a59ed572715b0e6ff7a3ad
1148 F20101129_AAAFBG fernandezdiaz_j_Page_052.txt
24a4c00a7df48cb88b93fd6f7250edac
03918695b21c50450729c106889801fd34cbecb7
24597 F20101129_AAAGDU fernandezdiaz_j_Page_050.QC.jpg
92398da95945cea2016cb13aa346783b
5e24db92620d4d7c1775f959c9a40a79c74c5baa
F20101129_AAAEWA fernandezdiaz_j_Page_101.tif
9d0dac372e5266a4224c6e64cbf5bbfa
f22aa4452d3dfcbf39a85871d8f65dd7f8e2ad31
F20101129_AAAFAS fernandezdiaz_j_Page_067.tif
2971bdd78b07165cecd7330ffab2e537
c4c5bb29500409cd495ece0c24c0a7210a09744a
2444 F20101129_AAAFZC fernandezdiaz_j_Page_086.txt
06c73c37c0519e17c64c907c2b75f768
eeeee2281eb8f8984f61b2aa0664982f524f167a
1707 F20101129_AAAFYO fernandezdiaz_j_Page_057.txt
3700e36f6f40824fe3ca1ef6827e0110
c044e2295aa71bd9b945e45db6286613b46d9edd
25605 F20101129_AAAGEJ fernandezdiaz_j_Page_069.QC.jpg
4d70f8ca92afdd9a1e843d7f79018823
82b6bfd479fad15d763cb6f28bd08221f83674b3
6912 F20101129_AAAGDV fernandezdiaz_j_Page_050thm.jpg
6553886637b5084b7563087b6e475472
d55cce08a6e4d17fd082f25b491ab1c9150e3e02
1051928 F20101129_AAAEWB fernandezdiaz_j_Page_093.jp2
73d43c5fa9532c6aaacca1f10a39b9e9
b4933f2b30b3742cd766c8a1e1ac0bcdb14c65d0
1051986 F20101129_AAAFAT fernandezdiaz_j_Page_071.jp2
d43ce280ef3cbb293af5fb0e6421f0da
3d1aaf46e98f10e70710e7fc20c19b582bbdaf02
1221 F20101129_AAAFZD fernandezdiaz_j_Page_087.txt
14d56765ddc93f471af41f3825d04c94
51fcfbfd0607344d22d440e05bc7b897b70b0b3d
870 F20101129_AAAFYP fernandezdiaz_j_Page_059.txt
0eadae313c9004377a0cd0d4cde4b719
83955e04960ac936324d997f479cc3a0c942addb
85083 F20101129_AAAFBH fernandezdiaz_j_Page_088.jpg
8677d24f930168004eb923f47fc24803
239aeff965b477d88d88f2bf0a7adb0088c1e3d5
7392 F20101129_AAAGEK fernandezdiaz_j_Page_070thm.jpg
8fe35d52aca3d233ea4f43f30e4efef4
46760f7e6a8321cd93260ab8049f37dd774c676b
25571 F20101129_AAAGDW fernandezdiaz_j_Page_052.QC.jpg
cdfdfb081c2e478cd528b0323ab65833
ce8943a83efd6afa20ab78640674707ddfcfef81
1051922 F20101129_AAAEWC fernandezdiaz_j_Page_091.jp2
ca96cd26699242eee5965523f4703c5f
558650e31c0a44e91fca3b07b177880935581d13
67276 F20101129_AAAFAU fernandezdiaz_j_Page_015.jp2
00657f44f0bb54fe5862a00f00a42e3c
2d13b6d061c4c63b25727602468d21098c9ef491
2674 F20101129_AAAFZE fernandezdiaz_j_Page_091.txt
f8df68b73fa10c98a9a2dfcf8b676634
acd3d41925588ca89a3a9eb9ee324e0281530626
1371 F20101129_AAAFYQ fernandezdiaz_j_Page_060.txt
7a6843ddb7d84ba3aa13ee1221932eaa
106ac9511cc83fb5ae93df97696aac5ae84e386a
F20101129_AAAFBI fernandezdiaz_j_Page_085.tif
bff42589f0f0fdddc1c478ef94d163c3
20de2216a7f3a07352b26a99e5acb9d2ceed8ff5
22022 F20101129_AAAGFA fernandezdiaz_j_Page_091.QC.jpg
64075ba2aba9b4a9d5c641ce207c504e
d9ccf68640c867efc475e3a685316d1b0a8cc20d
24387 F20101129_AAAGEL fernandezdiaz_j_Page_071.QC.jpg
f7fdec72da9a832eb75d09241c7fed06
94aa33f1568b28f97009aaab596833d54b623eae
25087 F20101129_AAAGDX fernandezdiaz_j_Page_053.QC.jpg
d6401d32adc3cc1e30dab9a7775fc625
8ee145ae46876425d242504ce1a74c3025f67739
22147 F20101129_AAAFAV fernandezdiaz_j_Page_115.jpg
8214486259a80a0a57d0446453a5c505
abbf92566cc61c6c927f7d6a5802a82e4dc5d884
1524 F20101129_AAAFZF fernandezdiaz_j_Page_092.txt
764c4fef960bdadd1c81f94091319ff6
4f331ee5bfe1a369e66534538e87eaf96f3efc99
1639 F20101129_AAAFYR fernandezdiaz_j_Page_063.txt
01f905d3c93e67d8a78f48f5d5a0b445
37e02f2cbbcf4ac0d46edbac5745fed6ab3c290e
F20101129_AAAFBJ fernandezdiaz_j_Page_091.tif
ffc388a60dc47042a0d195061b7b3e0b
a13880f1ddc2fde37c2d88efbd6f136e18c4fe9f
27524 F20101129_AAAEWD fernandezdiaz_j_Page_024.QC.jpg
ea8da7ab65472ae8168776ea75bfaf1a
11ac1a78d4fcb88b78bf6348bf1b9826ff392f25
17777 F20101129_AAAGFB fernandezdiaz_j_Page_093.QC.jpg
71f86a5c336acb5bb3323c7a2c3d5529
76559e749ae9d9aa27df2c9b0d542ad505c079c8
12457 F20101129_AAAGEM fernandezdiaz_j_Page_073.QC.jpg
02c2a57789349102ee13516513b85b29
33870df1eb6a4161e7b7ba11f9be95463fa4c2b1
24531 F20101129_AAAGDY fernandezdiaz_j_Page_054.QC.jpg
daae2e46dbf23e1a0d5f37fd1a781a1c
d77726fbe601991147dbb961ac3c9ccfa94bb96f
39910 F20101129_AAAFAW fernandezdiaz_j_Page_088.pro
c51c65abf89458291fb42e23bf612fa9
12b390380c69667448295ebe4a1fe7318d1d0ce4
958 F20101129_AAAFZG fernandezdiaz_j_Page_093.txt
05add40da03d7b2d531f18c8728f2c91
c60eb3430151b6774553740559b238bbb4461f26
226 F20101129_AAAFYS fernandezdiaz_j_Page_064.txt
5fde94c3b9276899fd53bbfe8aadcbfa
4ff0e6f2d1aca85fe0636ed0f785bdbbb39474ee
1739 F20101129_AAAFBK fernandezdiaz_j_Page_024.txt
ff5e182bb0971a400eb5d05621990385
3dfb27618285da45351fdc66b22034909833e4eb
6512 F20101129_AAAEWE fernandezdiaz_j_Page_080thm.jpg
f98bd0863bdd6d656f40e51f50d9ab66
84e291eca5583659fa8d06df6753e7af5e30779d
2670 F20101129_AAAGFC fernandezdiaz_j_Page_098thm.jpg
c97adfe97774172ce290da417cd8b846
84623d797d4c8719cfc77db1be345bbdcb297a2b
3741 F20101129_AAAGEN fernandezdiaz_j_Page_073thm.jpg
5228a4d4dc667d190bee2f2c4e575e65
15c89cc6f02ec2b6d19b503400dc1611b282366a
21070 F20101129_AAAGDZ fernandezdiaz_j_Page_055.QC.jpg
7a911f7b6c09b3720216d054bdfb9eed
a85961ddb2ac8c6d192dadde1218a56fbf11b406
93618 F20101129_AAAFAX fernandezdiaz_j_Page_005.pro
bb146add6f7467f5c7c1470bcf18b7b5
5800fe05c83146b96e5a7840191e4ac3fae9b23a
1413 F20101129_AAAFZH fernandezdiaz_j_Page_094.txt
a928dfdea20d8c704868c3e94db96dde
a8736eae8150deae90b1bc828c33a388dd73a1ee
1075 F20101129_AAAFYT fernandezdiaz_j_Page_065.txt
0706ea525f2412f8e4ffd06d2b6e064a
44e48f428b96b963a6b12db6d9659ebe69d37762
88761 F20101129_AAAFCA fernandezdiaz_j_Page_011.jpg
dcd9a202a2e0f26ccf57064f94763f78
ac70d4486e13edf78034a406ffdb3e3bd7fa2c74
120633 F20101129_AAAFBL fernandezdiaz_j_Page_132.jp2
249cf910375a4140001a134dd6fb3f4c
5b476e7625a64bb3dd5ec77ecfc3b943747c4633
5521 F20101129_AAAEWF fernandezdiaz_j_Page_103thm.jpg
7846166315df2559da1c7db117f080b5
867ff49e534223a7379250b2b62d64c64033b796
23285 F20101129_AAAGFD fernandezdiaz_j_Page_099.QC.jpg
6b7094f2633de5d433890ffcc606063d
50cf883a61c0a39ad00bbf4986b9a85c00ca5b87
24713 F20101129_AAAGEO fernandezdiaz_j_Page_074.QC.jpg
a3c59956a1a08996fe12bdf8086b6559
b9f5ae7df70a858a2ad453c2e938847dd56172ab
1847 F20101129_AAAFAY fernandezdiaz_j_Page_077.txt
ef27cfb6a854735dec9432f557b0388c
3af1d9dba37802d90becf78aa91b620552ad3c52
3307 F20101129_AAAFZI fernandezdiaz_j_Page_097.txt
93cb411230eedfe42781785686f5b6a0
329c2bd834be04bb081bfea8e62798e71316648d
2046 F20101129_AAAFYU fernandezdiaz_j_Page_067.txt
6ebe61b6a554cb7cb3723720c38db91e
034d5a0a260a47f980feb8cd0cf6cd8951eafdae
527 F20101129_AAAFCB fernandezdiaz_j_Page_111.txt
7e707ec331a335910603b41a7227175f
def6b2ecfc512f11ff5ec63fb2dc8304fcd8a6ee
6856 F20101129_AAAFBM fernandezdiaz_j_Page_025thm.jpg
1c4134493af5a0d82fdf90cd9ef7fbd3
8bf4701a9e7a893f12371ff48facd5380eeab5ad
F20101129_AAAEWG fernandezdiaz_j_Page_080.tif
6b5bb479ebbf24b43a62006b8ce311c2
b4d124a5debf768766d3ec67b2337ad058e42ad2
6329 F20101129_AAAGFE fernandezdiaz_j_Page_099thm.jpg
58c976f24cddc34c630b5a10e27cf0cf
800b46d51f9ade98284f10090f39e1c52efe16a0
6765 F20101129_AAAGEP fernandezdiaz_j_Page_075thm.jpg
945a51a5dd2e6a7a7298e69cec863bc9
6a9e212af6f5b08544660e4059a8f34cdfe65b9f
F20101129_AAAFAZ fernandezdiaz_j_Page_130.jp2
b3ab880167857abc94872aba09aede9e
1c5ac0d4bc8fb3e84165427fb6112d264f33e084
607 F20101129_AAAFZJ fernandezdiaz_j_Page_098.txt
67896c6caedbac0cb1e32e57226ea02a
1801f017f8472e6fa47eda10ba48d86caee9dae6
F20101129_AAAFYV fernandezdiaz_j_Page_069.txt
d9e99200c4fd7ae0ab207adba3cfd721
e0f63dcc17feb46c9bb6fc09e4a7a48dbe1b648e
22097 F20101129_AAAFCC fernandezdiaz_j_Page_010.QC.jpg
f5e1603b52b2cd4b7a40b5b53099d779
3a739dfd28176259ba0d070a31eecbc84cdfe3be
F20101129_AAAFBN fernandezdiaz_j_Page_032.tif
a0976d6d7365173dcfeaf29cf173383d
87f7d61ea4dd85cb3898e4b9bae90fd34f6763c8
8024 F20101129_AAAEWH fernandezdiaz_j_Page_066thm.jpg
50a6041356cb537460ef51e17c45f150
b407cfbaf4e4efcc40df97855575606589dd86cd
23307 F20101129_AAAGFF fernandezdiaz_j_Page_102.QC.jpg
015d0f40477d882dde8a0c6b9befd931
968b4c51f62558e4ba6974d7e8101443bc0dca17
6692 F20101129_AAAGEQ fernandezdiaz_j_Page_076thm.jpg
187b3254b37b58cc37c1b83f3c10874f
860050edc85712454914d5818f8dfbd0f1b78271
2022 F20101129_AAAFZK fernandezdiaz_j_Page_099.txt
e68f65ebb1e8b1066eb5e05613429e86
d3378fdf442135ce5dc19896474eaf3d3ab4e888
1735 F20101129_AAAFYW fernandezdiaz_j_Page_071.txt
81ed8f07071e349089b1fefee238a8c9
f91d7186d347321ce9f9513b62584481657a169e
1051958 F20101129_AAAFCD fernandezdiaz_j_Page_076.jp2
1bf49f5d894cb697c9b711a007f1267b
fa8b80acd271159922aec93e52c2fe09776bd173
7062157 F20101129_AAAFBO fernandezdiaz_j.pdf
d5f7f46bd2a43e43379caf1b84dac48a
a4724ef1575032a09572693ce35f6d31e20621aa
7034 F20101129_AAAEWI fernandezdiaz_j_Page_030thm.jpg
1d55d972fe0f20fb540bf84a850a147e
479595d5cae72f0d1ba8a0aa1ff170a310e8aa4c
24042 F20101129_AAAGFG fernandezdiaz_j_Page_105.QC.jpg
0efd3f3862c51760105a9aa2f05a0f7b
694e546addf4ba1dbe6e3232367e6cb31f7fbc58
24832 F20101129_AAAGER fernandezdiaz_j_Page_078.QC.jpg
6bb4242d1f9150f7d27527f0f8f88bf7
684b534e4cb737ddbc945a18c42d9635f658225e
2210 F20101129_AAAFZL fernandezdiaz_j_Page_101.txt
39e664024c33900ca4fd4e81f26e7d14
8267414cd22adecf00f5ea122884f84b314f554f
919 F20101129_AAAFYX fernandezdiaz_j_Page_073.txt
e8ed166f7879fae1c9ea227c78c7de74
fc0abfb8738fd9c7c238486d33a2431724f3d918
24893 F20101129_AAAFCE fernandezdiaz_j_Page_077.QC.jpg
567e53f43a13224966bc482f1074e538
8e5afe5e12d748e17e643ef0963cb0117e5f17ae
6018 F20101129_AAAFBP fernandezdiaz_j_Page_003.pro
e43cdc4b354a4a26336c866534fab612
990e9a520755cac46a4cc3eb29a1b1f4408dc356
25867 F20101129_AAAEWJ fernandezdiaz_j_Page_097.QC.jpg
b7c83ecee60a089fdd707a8c8ca20cdf
f78aa8d00a5247a125854852b48bbebc63f6ee2c
25448 F20101129_AAAGFH fernandezdiaz_j_Page_106.QC.jpg
aee528fd73a903b5e3109bfb30374b0f
3e3b6187763c2319ddd06974573d6074a05cc586
6025 F20101129_AAAGES fernandezdiaz_j_Page_081thm.jpg
677e626d18a329f3b021bc9557c083b5
bfc7550acdd67c0954050da7b0cb994406680da0
300 F20101129_AAAFZM fernandezdiaz_j_Page_106.txt
ac89a4d527cbdf746c974c7d06d71c09
dce848af77d7353bd85891c9197509eddd1e0958
2089 F20101129_AAAFYY fernandezdiaz_j_Page_074.txt
602354aae7451aef09850aaea8bbbef7
0bac7755b0f91bc8514143f3a12a284387ee8cab
52855 F20101129_AAAFCF fernandezdiaz_j_Page_073.jp2
ac02af097300face21fea46957f3ca0d
52c1f469e4982e5279046a7d5cb64267fba47a70
83291 F20101129_AAAFBQ fernandezdiaz_j_Page_052.jpg
775e0166f1af2c804cadb09c37b92d9b
dcae2210dd09290f66f37e40b92a9df362b3c2a2
69418 F20101129_AAAEWK fernandezdiaz_j_Page_041.jpg
cc430cc2d652c610d8071d3012cb877f
7b50121c336ad36a4d0d8f8fe4181041515f6c84
7382 F20101129_AAAGFI fernandezdiaz_j_Page_106thm.jpg
918aff312a564b9c2570688064e710cb
6afbc6089b9551e4aa13437d796a109492c13afe
20449 F20101129_AAAGET fernandezdiaz_j_Page_083.QC.jpg
8ca514bf8a968a66743d2f5820c2db77
5f29547609223d316db284fe006945b92987f14b
1361 F20101129_AAAFZN fernandezdiaz_j_Page_108.txt
9ff046dec5070429d1f01c293a5ff4a3
1c83aa04171cfe2f2b838330e27bd161a165ec58
2380 F20101129_AAAFYZ fernandezdiaz_j_Page_080.txt
a978f19a9cd83b35915c9d805ec64883
ab2a5481aeae2d6f1eca5e378af6e9116018267a
22952 F20101129_AAAFCG fernandezdiaz_j_Page_035.QC.jpg
b4d2a2f8d8b65962c7bb8d3838a5c92f
4ab97434e9f12243383c44ba5e8a1f5fda0937c5
71225 F20101129_AAAFBR fernandezdiaz_j_Page_099.jpg
f4d354209c9e1b588088faa7609aeef8
1ebbdd1a6d51f3d54ae15457cf3f3a9e4713b491
5813 F20101129_AAAEWL fernandezdiaz_j_Page_086thm.jpg
8d852fbac87a27f95e4eb05bf05cb937
ab2a80c7284f4a73e1428359770af9e569058f45
22488 F20101129_AAAGFJ fernandezdiaz_j_Page_110.QC.jpg
7c292e210e6fdc1a1deb36b0f26c2127
95e6729d1ce78fe101106efbca015f6e222635a0
6801 F20101129_AAAGEU fernandezdiaz_j_Page_084thm.jpg
9d909c9b925ebd73782965df3b290051
7ed8a86dcbb08a724249dff7da6987f4f5ae2c5e
6536 F20101129_AAAEVY fernandezdiaz_j_Page_118thm.jpg
0578b5d489891075be5beed62f043169
ba1648ae28da2caa7013e903d0fd67863d32d218
6196 F20101129_AAAFCH fernandezdiaz_j_Page_065thm.jpg
84b491899816aeb11458355f98539b58
dea552ac57af3c5448924ac5ef241b135a494682
1532 F20101129_AAAEXA fernandezdiaz_j_Page_041.txt
0508791626d88b0c04df0254abcda793
40f17ed0659bbec6fd7afe4e3f37c65287b79208
55459 F20101129_AAAFBS fernandezdiaz_j_Page_101.pro
2e98d5bec3e7e605a22359f2c7b39ac3
0ad8cdf7928ebd432e16b9193a60fb911a3e8b23
24911 F20101129_AAAEWM fernandezdiaz_j_Page_085.QC.jpg
a26017ce098fb4182aeb6183df5f5c54
5183f84f5d50e376735138f6fb88fc2436480312
993 F20101129_AAAFZO fernandezdiaz_j_Page_110.txt
c90df211b14772693ed4294048115360
aab22398463a7eacd1a880039dece8e83b1c5553
21254 F20101129_AAAGEV fernandezdiaz_j_Page_086.QC.jpg
12e0bd3d1651c5415512abba42201f78
10df198255e1d6371ffec90e049da451924258ea
2481 F20101129_AAAEVZ fernandezdiaz_j_Page_089.txt
9749672926d5a67a270991346a651a50
cc1eac82368df89c06b3b8597f8f9559f4644e85
1189 F20101129_AAAEXB fernandezdiaz_j_Page_121.txt
6a9142568984be2770806b186f792315
864ba040df74de7c5c080cd60f26d0dad42d506d
15481 F20101129_AAAFBT fernandezdiaz_j_Page_122.QC.jpg
42aaee1dafb6a764de6cd7ead963b398
889484c148a3dd583d29fcda8de00412532410cd
48727 F20101129_AAAEWN fernandezdiaz_j_Page_090.pro
f4bac92851d03bdac6940fdf4f98b2bd
613576f5ef5de2546eafe34158efa5df69716ee2
1289 F20101129_AAAFZP fernandezdiaz_j_Page_112.txt
43dd4220a53b8ea91a8c4aae6794c56d
970c07fdbb60d04e35019e3f7673c99e502b9638
21529 F20101129_AAAGFK fernandezdiaz_j_Page_111.QC.jpg
8e3c658f11f89eff286caabf642ec762
6409c92f12a1a4e32bb7dd82f7fe92b63cfc01c1
24213 F20101129_AAAGEW fernandezdiaz_j_Page_087.QC.jpg
12f5c5c3cfb56fccee34f7596ef68fae
3d97f9b233736112bc5b287d1fd39df60f66acf2
36525 F20101129_AAAFCI fernandezdiaz_j_Page_076.pro
42f644c9668a3b091f9f6d20bf96ff6c
ed805ff2ee7a683d5d2fd06771fdc9144d3cc368
76512 F20101129_AAAEXC fernandezdiaz_j_Page_021.jpg
c04c175eaf5ef2ba9574b73747e1b464
2e50a6b129f667e489c26f40c69b7bbba3ef7e98
1742 F20101129_AAAFBU fernandezdiaz_j_Page_031.txt
9d62a37a1da66e37a974c94e4ad20b88
bcc266ff59a0cc633db983660cae3714a3fe25fc
6738 F20101129_AAAEWO fernandezdiaz_j_Page_102thm.jpg
7d2438289780348c21369302cf4b3fa0
790360ccb1d772f24a27ce079bcf828e1b699f87
2414 F20101129_AAAFZQ fernandezdiaz_j_Page_113.txt
03172775c2b4eb174fcc1fc6364774f6
17e9ced28b8eeb06db5044eb63e9cb7848d1a1fd
19359 F20101129_AAAGFL fernandezdiaz_j_Page_112.QC.jpg
31af77fe43d49e1a7f661f05efa9c7e8
4e0a6030c8899727ca47ad0f9d18b3986c045e46
19527 F20101129_AAAGEX fernandezdiaz_j_Page_089.QC.jpg
b38858edb15739c1315b7f446412e2fe
e381563cd2069f81409f2b57e9cd313f7853eb3c
4602 F20101129_AAAFCJ fernandezdiaz_j_Page_122thm.jpg
fc61706a52c267eca82151d55c2566d3
b5d6cdd3544302edd0cdb13bdc6795c6b6957765
F20101129_AAAEXD fernandezdiaz_j_Page_123.tif
1c6896baf93dfc92f5e2be9ef6c42dcd
25362a9d8691b64d2d16d490e3919165c8318380
F20101129_AAAFBV fernandezdiaz_j_Page_105.tif
eaa0b5d85e62596e3dcdf7d04f18bfd3
f357e2ebe6a76e4c7ffa981ca70ba9965260f8c7
25722 F20101129_AAAEWP fernandezdiaz_j_Page_019.QC.jpg
ac0b853410b5c8a4050a5372c59a3963
5de3c845c96237dcf8367f438bd5396fa59663b0
444 F20101129_AAAFZR fernandezdiaz_j_Page_115.txt
3e7c60e08f1a13437d29447b0df817aa
b4714b08baf684fa2638def0335f623927fa3a49
5713 F20101129_AAAGFM fernandezdiaz_j_Page_112thm.jpg
3e5dac876d8f6bbf9522ccf21150ae68
ce69fcfcd6b455b12872ed84fe710cb68a31d7d3
5819 F20101129_AAAGEY fernandezdiaz_j_Page_089thm.jpg
ab1ef9fd14b6eb2f350d7f34c1479695
25b5501a43f2599b8e02c08cf69b515b893a94f1
4343 F20101129_AAAFCK fernandezdiaz_j_Page_006.txt
0b57d864a899ace9d1772dce34e17984
59ac4f541d8773cf3a9eced5e78a32261d39bfac
81767 F20101129_AAAEXE fernandezdiaz_j_Page_032.jpg
e8cfcc923445d0c70afca9565a9197b6
ddc03613860742efa7f28980cfaf3afa7a0c130d
52986 F20101129_AAAFBW fernandezdiaz_j_Page_019.pro
6f91ee13f692ce309f0d49270d53d76b
9550528667bc6496a04fd1bd166ab201714f3268
55202 F20101129_AAAEWQ fernandezdiaz_j_Page_030.pro
5788ed46994830df587818a255a29fb3
ae67982e2b4afcf65613ff6b84a33549b7b4fe63
2147 F20101129_AAAFZS fernandezdiaz_j_Page_116.txt
f024a822a1ef9f008b24621c9e48c6b5
90fba4a4ae54563365340fa1707974a6709f29ee
16845 F20101129_AAAGFN fernandezdiaz_j_Page_119.QC.jpg
f071f1834647864ea80be635e7062558
d889c04ad539a395f2f08ce0ef2f26b25c3f69b7
21699 F20101129_AAAGEZ fernandezdiaz_j_Page_090.QC.jpg
702b5dbc1bf862e202cf5a40a37d381d
40a7cbdb986c331fdf3744bd92fbb9757dec49ce
7556 F20101129_AAAFDA fernandezdiaz_j_Page_079thm.jpg
4a884412856dd5a41e9764421170ce7e
ad13be97a87c89741b0ed02a712e25c59d7eef64
F20101129_AAAFCL fernandezdiaz_j_Page_115.tif
510044ac9e687c856ffcb098a076a098
3635ed8bd19124b3eb215c911d42bf85dccbb183
55046 F20101129_AAAEXF fernandezdiaz_j_Page_033.jpg
e3ad42cf64e5966b56dbf2f144573101
36bfcae06b7ed5614d8c4e196760444987f67514
87620 F20101129_AAAFBX fernandezdiaz_j_Page_097.jpg
821e885e76833a9ea118ef0a8e004335
77140ac8c2403201f39e90552eb973743f776ea2
966806 F20101129_AAAEWR fernandezdiaz_j_Page_124.jp2
4493e2ef54449f78fb665e7d93d58d94
cea7faa5d43126fc553522d52bf9c8efa24b80a4
F20101129_AAAFZT fernandezdiaz_j_Page_117.txt
3313e4e1fdd88702ebe52ce26d4c655c
10a6033b10536c926fc39eef0cf01f7bca2c0c25
15559 F20101129_AAAGFO fernandezdiaz_j_Page_120.QC.jpg
3a3247bdcee6496e092527accf81d74e
298f917b8d097474380b76a3b9e21942a3bf5a06
F20101129_AAAFCM fernandezdiaz_j_Page_070.jp2
90ce4439bb995f14eed0532c2fc58930
4b95d1b1bb2e7a06d9b5bd91df3e877710836a80
F20101129_AAAEXG fernandezdiaz_j_Page_054.tif
c10161ddaaf25fc739b3fa7b07d5e790
5bb0c5d799d0f0935b744e024faab5551f86ca57
1182 F20101129_AAAFBY fernandezdiaz_j_Page_015.txt
18d44c5fdc0927aed250192564bc80bb
d0c92c74636c575c803e78f8c6b9b2ef3902c4cf
24221 F20101129_AAAEWS fernandezdiaz_j_Page_025.QC.jpg
43b89902c3284a26e4d4c5a8b1888017
0ad3ce2c30b721df42c6ee07af9f9f1604ae65a0
1170 F20101129_AAAFZU fernandezdiaz_j_Page_119.txt
f5e0d8ee7dd93620cbfe0e14ca8f6ef3
7ae46f4e8c7218c3b260cb64708534ede5d927c6
1051974 F20101129_AAAFDB fernandezdiaz_j_Page_069.jp2
74a2acb46cc85dca8ac750fbc60d876c
e98084c51d27461ff2e186d8d34abe5a6518830a
16301 F20101129_AAAGFP fernandezdiaz_j_Page_121.QC.jpg
58e41ff5de91e053869f2a47ebf311e5
f15d05603371b250963ec44122063fb212737be3
4695 F20101129_AAAFCN fernandezdiaz_j_Page_120thm.jpg
599d6e973ddf7798c2e1b3b7c66fde02
b1dcaa98feb93506076b005b2e74b708052ffe8b
15548 F20101129_AAAEXH fernandezdiaz_j_Page_008.pro
bebd5a922af307de905401a6d1c84859
ac87e5482295ddda400042dd2f4394e130439873
70760 F20101129_AAAFBZ fernandezdiaz_j_Page_060.jpg
1516681bd01d14b24e47dc363f0c755b
b342ee7f79e1dea55811f0eb5d5f334ea8128391
75953 F20101129_AAAEWT fernandezdiaz_j_Page_031.jpg
b40b2710e3a9c59a6162e96861d10ff0
9994db789e2c12e4990063e954af6a013d595b3f
1023 F20101129_AAAFZV fernandezdiaz_j_Page_120.txt
923f76493d1d522cb0666b271ed17666
664c09006e0237badbaf3419392c4412cb64d9d5
33769 F20101129_AAAFDC fernandezdiaz_j_Page_054.pro
a76db0fd20ce3ecde96adf6303d25ac5
1f8584587189ba5caa1dab23f8be719ba644d640
6146 F20101129_AAAGFQ fernandezdiaz_j_Page_125thm.jpg
f5bf203a6b72ac4d1f1621cca3acba08
ac5deb938b60a9686e7eb63e06c456564a0f9c8b
40741 F20101129_AAAFCO fernandezdiaz_j_Page_063.pro
06143f7015a55a130167cdf845cb4cf9
b76a5a0b75564870036ad381c22b08262612bfe3
25044 F20101129_AAAEXI fernandezdiaz_j_Page_018.QC.jpg
85585cbf6ca99c9985d352b27306e6db
a092c0702dd9a6ea607f63e52bc1b5b9161f4ee8
6713 F20101129_AAAEWU fernandezdiaz_j_Page_116thm.jpg
c672fae7bef5021d4ca7170c9205e892
553995937dab04e7036300fac0dc4b9e96fd81dc
943 F20101129_AAAFZW fernandezdiaz_j_Page_122.txt
e144bd7104010483413b9e0e0a414ba0
d27254dad83742a99a1955322f41a534ca195a32
107907 F20101129_AAAFDD fernandezdiaz_j_Page_102.jp2
a08d2a3d268ff57f206d9dcff525efad
82221363530a376dbc61cbd37afb834f8feb3d72
29938 F20101129_AAAGFR fernandezdiaz_j_Page_126.QC.jpg
e20ee78c97e2464b6d214196d8fd29e2
b603748442344a3fe512494b722214ebd0d352a7
F20101129_AAAFCP fernandezdiaz_j_Page_098.tif
a8cf60d08e292b25970de056fce12a14
44a3da84ae0fe132fefcc827ae6cf1d4b5c5e161
43274 F20101129_AAAEXJ fernandezdiaz_j_Page_024.pro
a481674d9419d0bea135930d7a3f11e1
d39b0d734d15ab009d8ece50d89b14f434c685e2
2449 F20101129_AAAEWV fernandezdiaz_j_Page_131thm.jpg
6ee387a8e2b484384f25aa29e45323d5
4679a4670bdc8e82354ca0029a01091072549862
1046 F20101129_AAAFZX fernandezdiaz_j_Page_123.txt
15583fc3b05f50210c39dce7b7a099b7
96a67d562519655626a258debd41e38c42f86bc4
44855 F20101129_AAAFDE fernandezdiaz_j_Page_133.jp2
a8b9e3e127e83fd16801b3e464522f82
dd496fb474b60b2a4c358db4b7e7253553548dc9
31554 F20101129_AAAGFS fernandezdiaz_j_Page_127.QC.jpg
5a680a879a2953e5e1071dcee25d9ed3
cb702bf27626567bba4b13b05566c5c33cb1119b
21044 F20101129_AAAFCQ fernandezdiaz_j_Page_082.QC.jpg
908b1936f72b217b0aeac52fd1346781
1c25a10bd70997b3513673d38cd2aee7ae8454d8
18668 F20101129_AAAEXK fernandezdiaz_j_Page_038.QC.jpg
1d1020e6475411b4631318c11178d094
b5be18700afb6a15c28111803363e92785be413b
F20101129_AAAEWW fernandezdiaz_j_Page_082.jp2
eedc266f7733fdaa8db44805b7700e4c
f0bcf32c161a055433679267ed4d54384217ee26
1656 F20101129_AAAFZY fernandezdiaz_j_Page_124.txt
38526e49d450719fa9acf3a60321fad0
d9e8970126809230b20904a31b540a5552aab09c
1154 F20101129_AAAFDF fernandezdiaz_j_Page_002.pro
266d733d4493d9f676b51fce336e6be7
12cfa685c4c108e76792c246052eb16c6d47a634
8020 F20101129_AAAGFT fernandezdiaz_j_Page_127thm.jpg
ec14ebba42aa27e9d369f8230067deed
8b4655145d2c513fcc2f6e13987da6d46d23c90a
6853 F20101129_AAAFCR fernandezdiaz_j_Page_045thm.jpg
26ed6395f954482ad2c6f8304de9e23c
051658c75591501e92a0ff6a9369baacf252b789
1510 F20101129_AAAEXL fernandezdiaz_j_Page_096.txt
3201a3052d22f3cc575a64ee82b997f2
981fd5a5b58fdab6cb8bec9aea6651563735ca61
50933 F20101129_AAAEWX fernandezdiaz_j_Page_047.pro
d5a84459bc475337904222b85ade962f
ee784a3f0f791259420218d1869b1bc0e224832f
3751 F20101129_AAAFZZ fernandezdiaz_j_Page_125.txt
1c9d05b88b08d5306fff0576742a2ce1
67154973764c73d0ef8b1715efdceb0623504af1
89387 F20101129_AAAFDG fernandezdiaz_j_Page_104.jpg
e77d1413fe37eb3830c19d91591bda57
6418c554afa2d54c88a7b34c18d6c693039a7ccb
7745 F20101129_AAAGFU fernandezdiaz_j_Page_128thm.jpg
0a367359df5aeff2aae0bcf0f02ef371
b58e75bda3cc2460862ba052550f555918ef0b6a
1051960 F20101129_AAAEYA fernandezdiaz_j_Page_127.jp2
2b96c14627715f9bcbef8addb6bd9f93
aec6cd8624bdacd9133d03c6a9e7d2e97906562d
66117 F20101129_AAAFCS fernandezdiaz_j_Page_081.jpg
6de8efc879bc887b1aff08cf112365fe
01ba930155a3828ba790f1d8b266a39612baa763
F20101129_AAAEXM fernandezdiaz_j_Page_079.jp2
197f7023ce2e921671bdb2bf49232a7a
cc6bf82443e9832501ac541d8b693ce35d44d936
7379 F20101129_AAAEWY fernandezdiaz_j_Page_115.QC.jpg
5b9b350559e6db7fde6f6a8186baf796
30ccc3a108503ee10f4a8ac71541092cf9c0847b
F20101129_AAAFDH fernandezdiaz_j_Page_062.tif
357f449f2e77aa51a3910872126d51f2
a07605a9ade6552fde9a70ca6932484e8ff7408a
30167 F20101129_AAAGFV fernandezdiaz_j_Page_129.QC.jpg
19f9c19475814b1b6a2dbee1a13e6fb9
cf43e0689a19ba591b405727ddbdb16c584ffddf
6628 F20101129_AAAEYB fernandezdiaz_j_Page_109thm.jpg
dc0795c71f376c2c90a6fc53bec6117b
b0f9609aca0210266f94268c20fc6cd7bdac210a
14086 F20101129_AAAFCT fernandezdiaz_j_Page_098.pro
e8df79d04a42a16a41c70588113fe70c
f50ce77f96b4647cfc682bfbba1a9088752503d6
43405 F20101129_AAAEXN fernandezdiaz_j_Page_013.jpg
97a02da5f234d54968fa4629976b52ce
9cf10ad5eac49047618c6c77aee07e1ffb3a4cd1
F20101129_AAAEWZ fernandezdiaz_j_Page_045.tif
3a74c39ae2a03ea1b4ab97137c1aabf8
de9ef9c673af4f476f333c6b564e5ef7c137cb51
F20101129_AAAFDI fernandezdiaz_j_Page_110.tif
a92cbff81c8e3c585fa2fb3020ada804
ff023862c7b000320d6ac6554f1a1666951447e9
30865 F20101129_AAAGFW fernandezdiaz_j_Page_130.QC.jpg
9112a9d337f4a7ac23ac06aa95a6af7c
9dde09a7d5264ca938805b0669ff70eecd782d2e
6852 F20101129_AAAEYC fernandezdiaz_j_Page_017thm.jpg
3b4541de6f39974cb4f548c602133a04
0d0fabb81461d5435c2088deee62f16133d0e403
F20101129_AAAFCU fernandezdiaz_j_Page_061.tif
052cad857b4272f8f4d24bc54f11d35d
9441919cdb6afd12e67bcb7bb9d0f3c349b78b72
F20101129_AAAEXO fernandezdiaz_j_Page_111.tif
d5a143cb80fadf62e12e462c98741d20
876f2ab4d6ee23c7f48483201c0c6dbfc35112c7
10463 F20101129_AAAGFX fernandezdiaz_j_Page_133.QC.jpg
ecb9381b9fc2dae128bed4f9dac0c36f
4aa4f8ddfb4271cced2663cc96fe0848aa4c9639
2173 F20101129_AAAEYD fernandezdiaz_j_Page_053.txt
51b8151f78b02cb13a2fb73f4fae3a07
f13ae679c5c534d670af113e90c6c6b92947ee3f
F20101129_AAAFCV fernandezdiaz_j_Page_111.jp2
b9dd6ebb642f03179a28f5d3e5af003e
f0ceff7ea6f0f8fa8af025276d61c9ce4ca0a530
4975 F20101129_AAAEXP fernandezdiaz_j_Page_003.QC.jpg
b49a1385cb6e9ba78443ff77d77f9b1c
f0aea279a587a502ec91896cc08ab89266fb4388
22324 F20101129_AAAFDJ fernandezdiaz_j_Page_009.QC.jpg
b673a4dbf7fdaa3cd806020f3c1a1abf
58261ebb2ee146b0580ababe075575ac358b7765
1282 F20101129_AAAEYE fernandezdiaz_j_Page_045.txt
c80e7e1372b8daf5dcac147694980084
1dc3c8d61c6c6d8cf0c76ac6759a1b94f5a57112
19965 F20101129_AAAFCW fernandezdiaz_j_Page_092.QC.jpg
1a83d24c554fc5fef318da4278af4acf
dc8bb130aa791187b007c133fde30ca444c75d47
F20101129_AAAEXQ fernandezdiaz_j_Page_121.tif
59f8fe0bdbcb1016c4551ae017eaa31f
6f89f0b6fa0f05419c929293e21d5cb14a366df1
26806 F20101129_AAAFDK fernandezdiaz_j_Page_121.pro
dbd547d47b28c1aa8b20f2664bf13d7a
c02c1e686990641934a3c9ee853adb1b6bc5a871
45145 F20101129_AAAEYF fernandezdiaz_j_Page_017.pro
3baa24192dff626e08f5a4aafa685876
539b2ef094db02131be87fb5032616752f34d72c
118273 F20101129_AAAFCX fernandezdiaz_j_Page_101.jp2
c1fb5031ac2fe205599f9c3c89f3d0b1
4685170267a8230fed2f3abe1e4769e681034ee1
47365 F20101129_AAAEXR fernandezdiaz_j_Page_035.pro
c1340007ba82893faf33b288b7c9e4c8
a60fa1ea74398845284e8bf736df3f0ff035edfc
32616 F20101129_AAAFEA fernandezdiaz_j_Page_133.jpg
2a11e0a0ee77388d5e3c53705e3a77dd
ca1cf37732023efa3ca6889a6e17f697d5373759
51737 F20101129_AAAFDL fernandezdiaz_j_Page_028.pro
1820edf46299c9cc3a159f11db89ab37
eb2481a3c132d1a3b787ac5ac160ad132dac0e7d
1051978 F20101129_AAAEYG fernandezdiaz_j_Page_049.jp2
062a0648a6a07d6427b4716073867cb9
e41f6278dceca123afd680d8575569a3e265e7e0
6825 F20101129_AAAFCY fernandezdiaz_j_Page_108thm.jpg
84fb1d0f86e34391032c5d9a23914194
428c53711b222ecbce0c9db7d54d6ba9e4cf2c62
252 F20101129_AAAEXS fernandezdiaz_j_Page_109.txt
78db30a0b29996bff9e7fb8f5a1e8709
e3b2badcddc778cc0b37d08e8da3142d0d2113ef
40896 F20101129_AAAFEB fernandezdiaz_j_Page_032.pro
c9f55823a54fa638c7165c2a44fb57cd
2f9dd675847eaaf2e97f4f89c81113da250df366
1985 F20101129_AAAFDM fernandezdiaz_j_Page_075.txt
e8a8331ffda7419e0e430fe414927356
714ec2371760ed1b19c18303aa95541dc52b3edf
77448 F20101129_AAAEYH fernandezdiaz_j_Page_044.jpg
323fd5bfa47658c02a6c08492b10e98b
80d73fd712ce6d96603ddca90a1cd4d15c16974d
25827 F20101129_AAAFCZ fernandezdiaz_j_Page_107.QC.jpg
abc1e79a78adc3d973952a9f527b45df
1cd7a624a86db6cededc4692d82587737721a1a6
6429 F20101129_AAAEXT fernandezdiaz_j_Page_105thm.jpg
7afdf4b1d289e4546af2a1740c86ee83
8fffd2a6bcea02b7df8fffc328cd18caa82bc187
1915 F20101129_AAAFEC fernandezdiaz_j_Page_114.txt
7bb765fff204174d04007a1727238a12
eb2898d33d5512c212a4cc8d48dd63b48fc3f458
77717 F20101129_AAAFDN fernandezdiaz_j_Page_019.jpg
d541e1669f1d100666310c5c667d06cd
b23b1f09621e47bfb9523e759641a77c8c154cc4
F20101129_AAAEYI fernandezdiaz_j_Page_076.tif
0827af0f054578aafe7d738f06265e2e
35c17aa9ae0687adb3275b18c53ad85da37c178e
2113 F20101129_AAAEXU fernandezdiaz_j_Page_019.txt
87b92b19fd867062f60a1ef6702260ae
6b0e53b82b0139d9143f7a3f39ea16f5db8e020f
2027 F20101129_AAAFED fernandezdiaz_j_Page_104.txt
6f48c4be842a8594c7f48616f3b07029
294d9d0fa77651cb1a223a07d64846353ba18018
2284 F20101129_AAAFDO fernandezdiaz_j_Page_001thm.jpg
4e10445bcab8c6268f7cf79db03d5558
18779832f632e11d6b94dd706f813cda1cdad23a
F20101129_AAAEYJ fernandezdiaz_j_Page_118.tif
11f7faa07bdc46405be692cbc390cecf
90ab3d6e7270ad4cb94ac0fe2a696947c1f6a8d8
587 F20101129_AAAEXV fernandezdiaz_j_Page_068.txt
be251a95490e0a6544241cbae12e381c
d12fc125b2c2434a15868055eaf6ba03a5e920f0
22573 F20101129_AAAFEE fernandezdiaz_j_Page_049.QC.jpg
63f0ee0c16371a8f2f63cc3d7f232f71
ab83d072150270814b1678a59bfaf7ed010da624
38505 F20101129_AAAFDP fernandezdiaz_j_Page_124.pro
bf5916e305c88e75e206a0dc1761e54d
97a7d88b5352fdc229031b2ffaf109006605c645
1523 F20101129_AAAEYK fernandezdiaz_j_Page_076.txt
37aefac3ae82c5c359dad7d9cdbd04a0
63324acee45c89a88f396a4df1450ad689741671
75941 F20101129_AAAEXW fernandezdiaz_j_Page_018.jpg
32a2b32221ae667bf9830c44ae3635de
4b1a1858fad7301b514bbeaa14330942cbb6aa2e
1051942 F20101129_AAAFEF fernandezdiaz_j_Page_017.jp2
cbfb6a2edecb2435bb2a8c0b5683b3d1
42e3160df529ac9c60ed1ea219861771ce90f8f9
1051957 F20101129_AAAFDQ fernandezdiaz_j_Page_042.jp2
de5dd1ce5f4d39f5202386cb67d007a2
d4fac9b3c772602c945e8dae493392c56212df38
F20101129_AAAEYL fernandezdiaz_j_Page_031.jp2
4c03319ba1677bd9c413bde2a49c8774
b6a9feb4365ea3f6937abc985e745a2a78eadb7c
6622 F20101129_AAAEXX fernandezdiaz_j_Page_074thm.jpg
82c01c405199321eec8c1aa5f3c5a985
ff66cf199a2d43ec4cc1e01372e6aa7c042cbd17
26492 F20101129_AAAFEG fernandezdiaz_j_Page_070.QC.jpg
5a58fdf55230528ac2fff2d865a68995
d30766f240f22c854cc943fe8863c5df4f5200ed
6631 F20101129_AAAFDR fernandezdiaz_j_Page_044thm.jpg
281acd532016502ab809f43dd1c5eedf
85b88e6f99e6bc3d0d5f3edc2e08e3b5c60dbd56
114809 F20101129_AAAEYM fernandezdiaz_j_Page_117.jp2
70a4057d72cc4cc3eb343a0ec2ec5641
8a4ccb077b4dec1c2bd96a2f2eb47160c860a5bd
20549 F20101129_AAAEXY fernandezdiaz_j_Page_103.QC.jpg
8f26e68ee5f989ddcddfbb6c0433136e
2b53fde38c64b38c417b567aa31367cd491cb420
F20101129_AAAFEH fernandezdiaz_j_Page_045.jp2
71536dcb5a536dc9dc60ae4af83be0fd
310f2b3e58273c6c8e3fe7ff87bc0058ae28fca8
F20101129_AAAEZA fernandezdiaz_j_Page_087.jp2
123b252fb2f416465e9e7b70ba64ec66
3cbf444c8a36c08401a897237825b6f6b489319b
118999 F20101129_AAAFDS fernandezdiaz_j_Page_084.jp2
38f4617b943ae7616b23be7eb550d4a2
4ef75997a8c0147fff0b8f9867f0ffe52886f582
48093 F20101129_AAAEYN fernandezdiaz_j_Page_058.pro
774ebf00bfeb39e631489d7fa549ff78
9439399a76c9bc790a662a48b8438a4ed62d6cc5
71985 F20101129_AAAEXZ fernandezdiaz_j_Page_075.jpg
9bf3c6c582c3d8dfc07ccb1f39240a60
50e7ef570abfb79a080547ce9ccb34c346f88826
F20101129_AAAFEI fernandezdiaz_j_Page_082.tif
e567205a8e68acb437d2d57d8da71cdd
61bfd2b447edc9e845ee8674fad7c90b3229fb7a
F20101129_AAAEZB fernandezdiaz_j_Page_013.txt
fdf0a415280e3a242dcfc9c2f90d7e29
8f48046560f5944d09d620ffcc9535d72c1af5f1
2322 F20101129_AAAFDT fernandezdiaz_j_Page_084.txt
e957f14461f5d6f256672e48b8eaa326
6cc6bb834e00a892c2c978c3ca59a1fe8e216115
113127 F20101129_AAAEYO fernandezdiaz_j_Page_023.jp2
cdb2e67c5757592f8f33425eb25e5902
81782fb816d5adf548c3d44e1cf20140b206bd16
117564 F20101129_AAAFEJ fernandezdiaz_j_Page_026.jp2
0d705f27415acb8384f5d4b15aa01b30
cbde535169241dbdb40646fa7a39a3e650883ac9
1813 F20101129_AAAEZC fernandezdiaz_j_Page_088.txt
e79b31b1730809310642730dd7b32a09
06a9e56efce089a369fc0a8af456440e12577dd7
F20101129_AAAFDU fernandezdiaz_j_Page_074.tif
2745786989beddfca678a16a02bea39b
4b7f74433c5897483ffedddd6a5cbe626a37b96c
F20101129_AAAEYP fernandezdiaz_j_Page_036.txt
8378d0e03e8dc3951a27ce6ba45a6db6
fc5288c57a53363404f6eeecb992287959652324
23749 F20101129_AAAEZD fernandezdiaz_j_Page_108.QC.jpg
94437d95bd4b1fe1599b5ced1cb601d5
9a89f59af3d0c970489fa565e115f9fd8576baf3
65919 F20101129_AAAFDV fernandezdiaz_j_Page_111.jpg
bd999502842efc0d08373ccff1213b8f
040eb027530b18e67c25837c9eb3f4374fe0a61d
69065 F20101129_AAAEYQ fernandezdiaz_j_Page_014.jpg
525ba83d7e2aed429d13d341b4540e48
0a36cef0c4e3950076c2946d0f3b315877bd2000
6280 F20101129_AAAFEK fernandezdiaz_j_Page_095thm.jpg
d3e93096ba47e5e8edd27f1134b3b961
ade17a99efacb44d874176044779fd91d70c6d88
2956 F20101129_AAAEZE fernandezdiaz_j_Page_090.txt
20a73ed731b460c875f7737d923045b7
cdb65365ee6a5b668f3161026d5b703b09cb9c49
2167 F20101129_AAAFDW fernandezdiaz_j_Page_030.txt
0d3f164542ca44456185b5dfbb2be4a5
342c74b76334b56be24d4d6a56c86b78c46c0fb7
6426 F20101129_AAAEYR fernandezdiaz_j_Page_094thm.jpg
f0762194e4180ad3a64f35c1fe322297
93dd346a3598291e5327a0349acbb2a46253cd73
8266 F20101129_AAAFFA fernandezdiaz_j_Page_001.pro
67313d22e1d073566f5840839a65f9b0
5e9b9e22927bac0e681bc947b1fb3fa9b438bbd0
2186 F20101129_AAAFEL fernandezdiaz_j_Page_132.txt
64c6d78119b3f36ba071e0fd22a178fc
34ee478f13942a0f4f86cbf25efcce70076476e5
82006 F20101129_AAAEZF fernandezdiaz_j_Page_068.jpg
673f31ef821f8a260c2cfaaf80994355
935e2138467474e935f590aae3b23795018328d7
32478 F20101129_AAAFDX fernandezdiaz_j_Page_112.pro
ea3bedf2c82e36a0ecfcd49a7af6c248
1cb91e21f23202d46df0a3fc0ea2db597ef52421
74686 F20101129_AAAEYS fernandezdiaz_j_Page_028.jpg
62c55116a5a5ded4d1e64033257b14d4
4743e15b933838e30c92d6b11f48556a43dc366c
80493 F20101129_AAAFFB fernandezdiaz_j_Page_045.jpg
ede7054242f82b2bc57b0db432d9e847
c1b174929df68b773957fcd09eee9b662b15160f
52221 F20101129_AAAFEM fernandezdiaz_j_Page_121.jpg
8dd0a391ee823db7a38bfc0de7262607
f94c72015dcc5c0ef50e65378f1fdf96904abc20
26363 F20101129_AAAEZG fernandezdiaz_j_Page_022.QC.jpg
ecc4e01b73dc0c00f458b8deb18b19da
979e2887f9891a30bd9a8d581eb15ae95e5e31f3
5150 F20101129_AAAFDY fernandezdiaz_j_Page_119thm.jpg
5c932c712b1bc381d777d302cb918d79
39e0c1f598a8b0f12a9527f6b8ce55efa8b72ce9
1051949 F20101129_AAAEYT fernandezdiaz_j_Page_010.jp2
e5cb2476ffebf3395d4de6fde082a6a5
3e525aaf3f35c775356411867f60b5991b6763ab
1917 F20101129_AAAFFC fernandezdiaz_j_Page_003thm.jpg
4659aff3c94b7722b54cadd0d3c4eea5
f543f64b0d2c73d6895c7744813776958448cc61
2012 F20101129_AAAFEN fernandezdiaz_j_Page_102.txt
335b95d3edcd9be63358b2c3c51f305a
305f276ef6dfa1dd4f9283fa556f3557b2435048
716158 F20101129_AAAEZH fernandezdiaz_j_Page_123.jp2
4ebcd052c16a7f79b0568f1c1d89f4f2
fdb3c3b8ea52a1e1995d67464cccba041851c0ff
2727 F20101129_AAAFDZ fernandezdiaz_j_Page_082.txt
40331b35a93732d218bd169ea70a1fd3
5170a8ea3d16e7a70da663e38df8bd3f39f2d21d
64747 F20101129_AAAEYU fernandezdiaz_j_Page_065.jpg
f2c0197d84cd666e76e579338d240681
8016dff4c201e8573eb95e0995837ecf01b2c94a
107524 F20101129_AAAFFD fernandezdiaz_j_Page_075.jp2
347cb7057ec76ac1db912ad3b6b70fec
d7ed29837703da8017f2a46a368d15795490973a
65968 F20101129_AAAFEO fernandezdiaz_j_Page_124.jpg
db47dd9f20c220d14161715f6035ba4b
35f2e3cedfb267d3695deb0a97ff60111b4a10d8
44227 F20101129_AAAEZI fernandezdiaz_j_Page_104.pro
efb04a20de9d5fd7bd4695018de71012
0b8c88788abe701d12b68618085c3fa28372deff
81526 F20101129_AAAEYV fernandezdiaz_j_Page_076.jpg
ae3a08cfda76306c4d18e977dd684a07
87c24f3bef646e8150fc1dc0e934fb09970f357e
40549 F20101129_AAAFFE fernandezdiaz_j_Page_025.pro
63d5309e48894c2170403cb58985f902
b567da4988161e8889407ddda2925e65f0efde37
78014 F20101129_AAAFEP fernandezdiaz_j_Page_009.jpg
c04317e26225ad4fad169e84ea5c4e03
9830025f2171ac0b347c6b4f8ec34d39661c3aa4
F20101129_AAAEZJ fernandezdiaz_j_Page_058.tif
c7aa6b196871af70fdc862eca02f8795
8bc5e9a8ed4b7edf5496be2beaca9c7ebef677f6
1051936 F20101129_AAAEYW fernandezdiaz_j_Page_088.jp2
ead63d940d6dc2de16187a54d3a7c849
98bb61e1b6cf1fc802fb404804b846b354fefacc
409 F20101129_AAAFFF fernandezdiaz_j_Page_107.txt
d9e2e3f00deb0f1a2e2529b55ef9461a
f2eaf38183763a3b3a7a26ae24fbf3e786bcd45e
47054 F20101129_AAAFEQ fernandezdiaz_j_Page_089.pro
f1d9d4fb30e57afb4a3145ee4930061c
f5c71ce8f7daa1e6debd5a2d6360657097a69df0
81139 F20101129_AAAEZK fernandezdiaz_j_Page_017.jpg
a3388b72cb1d614c1bdbac9de713ae25
33928bf97bf12c4e5264d9cb61459eb99c02999e
64036 F20101129_AAAEYX fernandezdiaz_j_Page_086.jpg
01a6c0c6a8a290a7252fe1c5fc4898fb
f1b0558a918f2991271a5d7461ca67a7ef1a7e6c
F20101129_AAAFFG fernandezdiaz_j_Page_034.tif
bf4855c9311c07445db0e647e01519b9
fa741339d6ac54aea34e0a17c249901183a630a1
20372 F20101129_AAAFER fernandezdiaz_j_Page_081.QC.jpg
070272ba1a1c345a3728005ed692468c
9297b34d4756e08fb4e4dc05bdc3c48075dcc899
1167 F20101129_AAAEZL fernandezdiaz_j_Page_103.txt
3b5c618aae9fca84a15fdadab558339f
385762b38498da75cf208783b9b390e9ea67143a
87984 F20101129_AAAEYY fernandezdiaz_j_Page_089.jp2
0d06b376ded6b90e9fcf26a2b2bf3467
3114851eb5dcb9c2e21986e81185b016fd42bfff
114812 F20101129_AAAFFH fernandezdiaz_j_Page_067.jp2
cc7fb6b189a28e76e90fa5811d7140f7
53bac0531d927f23438afcb6e2788e64e60bd5b1
7578 F20101129_AAAFES fernandezdiaz_j_Page_059thm.jpg
e248c589a1c0a2ea3b9781d59c29b6bc
7050f8dadc6c6cbbb29ef76550c8a043dbc3822c
F20101129_AAAEZM fernandezdiaz_j_Page_032.jp2
3029446a79a68af7012fe9dc4bbc66e7
383f571c358e017616f58c2f4d7a523380a5e4fc
51478 F20101129_AAAEYZ fernandezdiaz_j_Page_026.pro
28891bf726ded118fb95d14af874b463
8ed8e7de47023e79a5c50b5d0f6487978961a04f
1069 F20101129_AAAFFI fernandezdiaz_j_Page_004.txt
5b50751e8576aa806141da363fba6fab
f135084ca983d03c9446449e0621d18509d6497d
6305 F20101129_AAAFET fernandezdiaz_j_Page_035thm.jpg
c239c8ce979a84073222c3a2a12bacf6
e6a15da89c3c8b69394c21bcdcd43bf2d8321055
5460 F20101129_AAAEZN fernandezdiaz_j_Page_093thm.jpg
35fd7270aeb4a7ba9aabfafe308fb326
38459ba6310534838148f86c527b69d4d264e8a2
7358 F20101129_AAAFFJ fernandezdiaz_j_Page_069thm.jpg
506963bef98ffa7e40b2f1709f62535b
036482abb1789a76bfa7df22a582bebe5d5bc149
F20101129_AAAFEU fernandezdiaz_j_Page_048.tif
16175e85632c2e3dd0ade0e8b7f9b7db
c5514f3967029751295be7103d267c8d02cf64d6
7297 F20101129_AAAEZO fernandezdiaz_j_Page_027thm.jpg
201647dd4b00dd595a2afe90e14556ec
90453bdb2ab5a6e33fac9b3b78ef48894359588a
62912 F20101129_AAAFFK fernandezdiaz_j_Page_055.jpg
d231032e6ca6fe8e3ccda3b8a16ed18b
c30b3e69fcf55707e6e0fdb4d1f219bfe677576e
24943 F20101129_AAAFEV fernandezdiaz_j_Page_076.QC.jpg
d5ccd74e7695c6c2df16d17b746602fa
24daa216b8dfe2c78909ae5cc570da1d0e9a3d85
F20101129_AAAEZP fernandezdiaz_j_Page_059.jp2
3128926edfae1513e6b2ea193e96503d
36811ffdbac225e5a853b2184e627ca6d5a92740
1051981 F20101129_AAAFEW fernandezdiaz_j_Page_065.jp2
296bdb3fd01d8a66706ebbc4f790c561
481f0ca06a84b8fac9f43ae27e96ea6282ae8fc2
68690 F20101129_AAAEZQ fernandezdiaz_j_Page_110.jpg
5ef78b884da5d40d137b04b2a8bc3406
32cb403054619a6331b92a68d12e1ad7551874a3
84823 F20101129_AAAFFL fernandezdiaz_j_Page_036.jpg
99d697801844e467c3cbf17fdd14e5a3
ae311b579228656d0895354dd076533ee97e94d6
26212 F20101129_AAAFEX fernandezdiaz_j_Page_132.QC.jpg
305015a2813b61f6330e8010459118e5
86a598ec073c2a5445dbc5eb1ac005dbc174075b
23024 F20101129_AAAEZR fernandezdiaz_j_Page_073.pro
be572e265e66f3f200f9225882c1ddee
2c09c5b2c93191c8fd7b292f3a25e5d8f4faa06f
2117 F20101129_AAAFGA fernandezdiaz_j_Page_018.txt
c13826b6abf121a1341962efac5d55b8
c49cf539f84e63935dd880e95120123133a46850
6542 F20101129_AAAFFM fernandezdiaz_j_Page_055thm.jpg
71a4d45ac17bebc894c82c21c1668b60
6c054e45b70978946932ae65d5e07c4ad2f2a19e
88789 F20101129_AAAFEY fernandezdiaz_j_Page_114.jpg
b31498f545d933aa6c8a04571881d1ce
1835fab90d4932acdbd200b7bc525f501dca3491
F20101129_AAAEZS fernandezdiaz_j_Page_041.tif
1f59823ca25db2e12a71644f5f2674d0
9d9e743a20d5d89e62a1962db33b202f1abd88bf
F20101129_AAAFGB fernandezdiaz_j_Page_006.jp2
6c242e4e2e4f0267717ebf985d9c828d
59f12b253d360f9db8dd97f646f2958c2381e767
24547 F20101129_AAAFFN fernandezdiaz_j_Page_068.QC.jpg
39f1b5922d971bae3c482b6f6bfb2fb1
c72de1f082e1bbeabb0394a96a686d2422b7f8f0
112853 F20101129_AAAFEZ fernandezdiaz_j_Page_016.jp2
4547c7432e456b3b557a0993a12cb142
f2316dde95d134d9c1adf33ffefda1bed795c711
28008 F20101129_AAAEZT fernandezdiaz_j_Page_103.pro
c37806338c5a40b2c6160edaadc462be
f3545c8b7591dd72adb9c82d48b70736f39cafac
6936 F20101129_AAAFGC fernandezdiaz_j_Page_053thm.jpg
dd2020affec49d5e4f81beca6a0825b1
2ecdbe21e171e79ea630f0c8fca6734ea6fcb87d
19991 F20101129_AAAFFO fernandezdiaz_j_Page_065.QC.jpg
08fadbeb4fee47f58291a545da780936
881f8fa2138024d1ba9b40e4f626922af1a26d15
6261 F20101129_AAAEZU fernandezdiaz_j_Page_110thm.jpg
8e62d08a13e8c30f16f5a6b228d29117
1dce3217bc4f547e1367e9e0d09e8fa5193ef0a3
F20101129_AAAFGD fernandezdiaz_j_Page_007.jp2
0aae2de9607c42dac267b9cc1dae1255
75a32db1d64649270782cf262d0f753777bcde99
72073 F20101129_AAAFFP fernandezdiaz_j_Page_102.jpg
7107ed18a7cf845029e7856b711fcd12
0b03e8c754974d9e388e5818fd3601797edbf6e0
25919 F20101129_AAAEZV fernandezdiaz_j_Page_030.QC.jpg
0b444fdd9c2ff0564d1ed751a807017b
69c447df87e1ad2f139eddb59227e912d162b2db
76240 F20101129_AAAFGE fernandezdiaz_j_Page_053.jpg
2f386bd3f9e8f2a4650c60a29dacd294
c707041fc4bd477ca4f7df0ef76adfaa7ece903f
23358 F20101129_AAAFFQ fernandezdiaz_j_Page_120.pro
3fcaec5de5843ec4f0f2b45c514893af
6895c5cbe7ddfbc6219553fc05323cfdd86987e5
7645 F20101129_AAAEZW fernandezdiaz_j_Page_088thm.jpg
77df07a15afe42a34b42ab7c5ac40da6
012b2470a10654dbf8eb6aa064abfaec80d73a0b
6497 F20101129_AAAFGF fernandezdiaz_j_Page_096thm.jpg
3581aec6367f0cad35a5395d5cf5a4d8
172d57eb5f9e54043bbc76f10ff148b043545c1b
F20101129_AAAFFR fernandezdiaz_j_Page_122.tif
148cdd681f5bfd1ddfcc5567dbd046c2
3dcc98258766945aa087e85d236007a3f64f6aaf
26689 F20101129_AAAEZX fernandezdiaz_j_Page_088.QC.jpg
57d58a5112abb188f70ed47adc5c73f1
ffd94c69308cea0a652faa289839c149ba0daea8
114395 F20101129_AAAFGG fernandezdiaz_j_Page_021.jp2
88932c0ab2fbd16595a63c6c8dd1e071
0839f5e76739a2417560c24429b249d3c6f7b414
2611 F20101129_AAAFFS fernandezdiaz_j_Page_128.txt
c46978b136fd15873b0036a621f5b21c
4fc1e3a752d8619ade148f01cd4da8a616b6c7fb
633 F20101129_AAAEZY fernandezdiaz_j_Page_048.txt
f1e632b24db18241657b63e44ebe4ad2
3e5a9d5573910d08fc69a93e0a58dd3c9a075050
19178 F20101129_AAAFGH fernandezdiaz_j_Page_124.QC.jpg
04047b7b25410a9e432409c532734616
00a3446cb641b9715f7155f7e8126b3784cec57e
F20101129_AAAFFT fernandezdiaz_j_Page_094.tif
2c955735590680b42917274cdd235f43
c6c9da5b8714259fa41387fbf984cf2591986e49
74285 F20101129_AAAEZZ fernandezdiaz_j_Page_050.jpg
2b17eb6f1ab462b0b1090df6e60cf1f4
1f3a2c2af89677e2bcadc07100cec7995da3ee1f
25224 F20101129_AAAFGI fernandezdiaz_j_Page_117.QC.jpg
5dcd165f6104d6d1e15161d8f8c2d011
22db9d59351cc1afe67f8cfecf646908ee87a7fe
F20101129_AAAFFU fernandezdiaz_j_Page_017.tif
6f8fa90c5415b5a49025237c9122533c
5976682fd89ef58a9b7644d43f328d81b7a26fc5
49578 F20101129_AAAFGJ fernandezdiaz_j_Page_120.jpg
e38cbfdcb3328a858806605b88c796a7
db34349b8973fddb7a6497bf365f54c2bb8929e5
6440 F20101129_AAAFFV fernandezdiaz_j_Page_031thm.jpg
d94f4b29e7271088954a22a385264139
abe1e7f07b752e35edb3359ef330dc04406b7d32
7012 F20101129_AAAFGK fernandezdiaz_j_Page_077thm.jpg
f29871d9127a11147e3e4f88e5cf4ff8
e169b8843ae06978200e630ecba1819ea0583e5c
2801 F20101129_AAAFFW fernandezdiaz_j_Page_095.txt
e7fd5fdde93d544f4b65a8a3aee3f482
fdbe4527c3304d88adc27997490f97eacd6f71ae
1051946 F20101129_AAAFGL fernandezdiaz_j_Page_066.jp2
25bc66953c47d2e34a7a4b919fc156b9
b6d031b700fdb2c8f7b832324775f3e63c749751
5236 F20101129_AAAFFX fernandezdiaz_j_Page_038thm.jpg
9e2ed5bdfcee460dbd4acaf143982539
27cc1e59bf0162242b6cbe480ece8a053666804c
126372 F20101129_AAAFHA fernandezdiaz_j_Page_036.jp2
0889698912c19e2281206a7f0b72ee27
8136295d06bd1d8ffae278c6eeb0cdc78fdb77a2
55913 F20101129_AAAFFY fernandezdiaz_j_Page_084.pro
0db487d96b043fa436a47d039fc87434
b52512d2d6bdc43c05ef840c0a58d0c5db7de71a
1623 F20101129_AAAFHB fernandezdiaz_j_Page_025.txt
6db4f6aba61f11e25de7b0969ec6a842
55e95f2c911941c002bf79c5c1f8caa5005e1a69
6673 F20101129_AAAFGM fernandezdiaz_j_Page_117thm.jpg
5f43ee7ea671832dae09a96e8685fd03
34783469325930fb4c99292bf4064de39b612b1b
F20101129_AAAFHC fernandezdiaz_j_Page_072.tif
58a78f0ee6c2e07f3890492329e880b3
f435022ec9263de4a9b9f9ec44fcb9c7b7de61d0
108643 F20101129_AAAFGN fernandezdiaz_j_Page_047.jp2
c0efe6c34e0517795d6d4b7450c419e8
a23c6ee6d73a8be6e21766bfe780caf6f4c6fb3c
63879 F20101129_AAAFFZ fernandezdiaz_j_Page_089.jpg
bd4e8d12521969f71700d51cde89d158
e9b529f78dc4ad0612f47bc864226be0ae17de93
1947 F20101129_AAAFHD fernandezdiaz_j_Page_118.txt
1355ad6c1d6dd3ecab0d00a77b602dbc
8e1b3fb5d2cdacab4f6348a18ba0476dc6c2e67c
48595 F20101129_AAAFGO fernandezdiaz_j_Page_099.pro
ec4857eb9b370015fb793fa5a32e3be9
1f8711d275c04a21c3e3bfa2deed41615473241e
548 F20101129_AAAFHE fernandezdiaz_j_Page_131.txt
21a9c9043f9403f1611dd1d740a42d3f
a9753ed1afd6bdbeed6c52ae3d83bc45e5f9b981
2244 F20101129_AAAFGP fernandezdiaz_j_Page_100.txt
c1173039d31b1b68233e4662d48cb51b
9aeb6f00e50b70a0e4f5e508dec57d26d1933cc3
7100 F20101129_AAAFHF fernandezdiaz_j_Page_020thm.jpg
86e6ddcd13e017d5743dba207d2bba22
56bef1d861d0cb3e66229466fb7bb85a8212795a
105870 F20101129_AAAFGQ fernandezdiaz_j_Page_099.jp2
d60cda43fbe3adaf0c8f9faa89d1ac45
cf10c3689fe585338ea42126358c75a08b747a25
7470 F20101129_AAAFHG fernandezdiaz_j_Page_072thm.jpg
26bc4374407a73a52ae108619d37c425
a527be43a366c007f565c2b92d12f1794459d1fd
4114 F20101129_AAAFGR fernandezdiaz_j_Page_004thm.jpg
7292c5718e014bf42c3afdd714e2e34c
9ec7719cbb0a18943242bbeff4b554df9bdf42ca
80995 F20101129_AAAFHH fernandezdiaz_j_Page_087.jpg
bfbf6283175f2a09da0d34e7b38de45f
43a706b4f1911911e701feb9cb21a25c08a36614
86891 F20101129_AAAFGS fernandezdiaz_j_Page_051.jpg
9a74835c0bc2337047f2d7d6c09b8c82
70fc8688e52c0000548cbd95c66e506fd9155f25







SCIENTIFIC APPLICATIONS OF THE MOBILE TERRESTRIAL LASER SCANNER
(M-TLS) SYSTEM






















By

JUAN CARLOS FERNANDEZ DIAZ


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA
2007

































2007 Juan Carlos Fernandez Diaz

































To my parents and to my siblings. Who I am and what I have achieved is a result of your
dedication, hard work and motivation. Also, to all the close friends who believed in me,
encouraged me, and provided their support.









ACKNOWLEDGMENTS

First, DEO GRATIAS. Second, I want to acknowledge the great support and guidance

from Ramesh Shrestha, William Carter and K. Clint Slatton. I thank them for sharing with me

their knowledge, experience, and spirit, and for their continued motivation and commitment to

make the geosensing graduates leaders in the field.

I gratefully acknowledge the persons and institutions that make the J.William Fulbright

Foreign Scholarship program a reality. Through them I had the unique opportunity to achieve

higher education in one of the top universities of the United States of America.

To all my friends, classmates and colleagues from the Geosensing Systems Engineering

program and the NSF supported National Center for Airborne Laser Mapping, who provided me

with invaluable support to complete this work.

Finally to the countless persons and institutions who in some way or another have

contributed to my success in the UF Geosensing Graduate Program.









TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ..............................................................................................................4

L IS T O F T A B L E S ................................................................................. 9

LIST OF FIGURES .................................. .. .... ..... ................. 10

A B ST R A C T .................................................14........

1 INTRODUCTION TO TERRESTRIAL LASER SCANNING...........................................16

1.1 G eodesy and the N eed for M easurem ents ........................................... .....................16
1.2 History of LASER EDM and Scanners ....................... ................. 19
1.3 TLS Subsystem s and Principles of Operation ..................................... ............... 29
1.3.1 LIDAR Ranging Principles ..................... .............................. 29
1.3.1.1 Phase difference m easurem ent (PD) .................................. ............... 30
1.3.1.2 Tim e of flight (T O F) ................................................... ....................32
1.3.1.3 O ptical triangulation (O T)................................... .................................... 33
1.3.2 Scanning M echanisms .................. ....................................... ............... 34
1.4 Technical Characteristics and Specifications of TLS.................................................35
1.4 .1 L A SE R T ype ...................................................................... 35
1.4.2 LA SER Class .................. ............................... ........ ... .......... 36
1.4.3 L A SER B eam D ivergence........................................................................ .. .... 36
1.4.4 Point Spacing.................................................... 37
1 .4 .5 R a n g e .........................................................................................3 7
1.4.6 Range Resolution............................................. 37
1.4.7 Precision and A accuracy ......................................................... 37

2 UF MOBILE TERRESTRIAL LASER SCANNING (M-TLS) SYSTEM ........................39

2.1 Evolution of the M -TL S Concept ................... ...................3...................9.....
2 .2 M -T L S Sub sy stem s ................................................................4 1
2.2.1 LIDAR Unit .......... .. ............ .... ...............41
2.2.2 Vehicle .......................... ........................44
2 .2 .3 L ift ........................................................................................................4 5
2 .2 .4 P ow er Su b sy stem ........................................................................................... 4 5
2 .2 .5 P an T ilt B ase ........................... ......... .. ............... .................. ... 46
2.2.6 Video Cam era ....... .. .... ............. .................. .... .... .. ............ 47
2 .2 .7 O n B board P C ..................................................................................... .47
2 .2 .8 G P S .................................................................... ..... 4 7
2 .2 .9 T iltm eter .................................................................................................. 4 9
2.2.10 IN S ................................................. ........ ...................... .. 49









3 WORKFLOW OF M-TLS OPERATIONS.................... .. ................. ............... 50

3 .1 D ata C o lle ctio n ........................ .. ............... .. .......................................................... 5 0
3.2 Data Parsing ... ..................... .... ...................................... 51
3.3 Data M manipulation and Information Extraction .................................. ............... 53
3.3.1 Visualization ............... ..............................53
3.3.2 Single Point Selection................................................. .............................. 54
3.3.3 M easurem ents .................... .......... ................. ...... .... .. .............. ... 54
3 .3 .4 P rim itiv e F hitting ........... .............................................. ............ .......... ....... 54
3.3.5 G enerating Cross Sections...................... .............. .................... ............... 55
3.3.6 Transform ations...................... ............. .............. ...... .............. .. 56
3.3.6.1 R stations and translations ........................................ ....... ............... 56
3 .3 .6 .2 C dropping ....................................56.............................
3 .3 .6 .3 M erg in g ...............................................................5 6
3.3.6.4 Geo-referencing.................... ...... .................. .... 57
3.3.7 Segmentation, Classification and Filtering............................... ............... 58
3.3.7.1 Segm entation ...................................... .... ............... .............. .. ... 58
3.3.7.2 C classification ............. ......................... .................. ......... .....58
3 .3 .7 .3 F ilterin g ...............................................................5 8
3.3.8 Gridding.................................................58
3.3.9 A advanced M them atical O operations ........................................ .....................59

4 TESTED APPLICATIONS OF M-TLS ..................... ............................................... 61

4.1 Common Applications of TLS............................ ................. ....................... 61
4 .2 P aleo n to lo g y .............................................................................6 2
4.3 Structural Geology .......................................................... ............ 65
4.4 W wildlife M anagem ent Conservation........................................ ........................... 69
4.5 Coastal M orphology ............................................ .. .. ........... ......... 69
4 .6 S o il S cien ce ................................................................7 1
4.7 Forestry ............... .... ......... ..............................................72

5 ST. AUGUSTINE BEACH EROSION HOT SPOT MAPPING................ ..................74

5 .1 M o tiv atio n ................................................................................................................. 7 4
5.2 U se of LID AR Technology ....... .......................................................... ................. 75
5.3 Data Collection ...................................... ............. ................ ......... 75
5.4 Data Processing ................................... .. .......... ........... ... 76
5.5 Results ......... .......................................... 79
5.5.1 Elevation Changes ................................ ........ ........ ... ................. 79
5.5.2 Volume Changes ................. ......... ....... ... ... ........ ........ ....80
5.5.3 Beach Line and Crest of Berm Extraction From the Grids ................................80
5.5.4 Across Beach Profile Extraction .................................................. 81
5.6 Comparison Between Traditional Methods and M-TLS .........................................83






6









6 SOIL ROUGHNESS METRICS DETERMINATION........................................................84

6 .1 M o tiv atio n .................................................................................................................. 8 4
6.2 U se of LID AR Technology ...................... ....... .....................................................85
6.3 D ata C collection ........... ... ................................................... ...........................86
6.4 D ata P processing .................................................................. 87
6.5 Results ................................................. 89
6.5.1 Simulated Profiling Results ........................ ......................... 89
6.5.2 Extension of the 2D Formulas for a 3D Surface. ......................... ...........92
6.5.3 Comparisons of Roughness Metrics From Profiles vs. Full Surface ....................95
6.5.4 Distribution Functions of the 3D Correlations Lengths. .................. ................ 95
6.6 Comparison with the Traditional Meshboard Method......................... ............... 96
6 .7 C o n c lu sio n s ....................... ...... ............................................................................ 9 8

7 FORESTRY METRICS APPLICATIONS .....................................................99

7 .1 M o tiv atio n .................................................................................................................. 9 9
7.2 Use of LIDAR Technology ...... .......... ................... ......... ........... 100
7.3 Previous W orks.................................................... 101
7 .4 D ata C collection ......... .... ..... ......... ............................................102
7 .5 D ata P ro cessin g ........................................................................................104
7.6 Results ........... ......... ......... ........................... ............... 108
7 .6 .1 S te m D en sity ................................................................................................. 10 8
7.6.2 Stem Location ........................ ..... ....... ...... .............110
7.6.3 Stem Diameter at Breast Height DBH ............... .............. ...................110
7 .6 .4 T ree H eig h t ........ ............. ....................... .................................................1 1 1
7.6.5 Stem V olum e ............ ......... .............................. ... ............ 112
7.6.6 Tree Biomass Estimation ................................................ .............. 113
7.7 Comparison with Traditional Methods ................ ............. ........ 113
7.8 C onclusions............... .......................................... ............ .. ....... ...... .. ... 114

8 SUMMARY ........................................... ...............116

8.1 C onclusions........ ...... .................................. .. .. ...... ............ 116
8.2 R ecom m endatio n s............................................................................................. 117

APPENDIX COMPARISON OF TERRESTRIAL LASER SCANNERS .............................119

A 1 O ptech IL R IS 3D ............................. ........................ .... .................................. .. 1 19
A .2 Leica H D S3000 ............... .............................. ............. .......... ... ........ 119
A .3 Leica H D S4500 25 & 53m .................................................. .............................. 120
A .4 R IE G L L M S-Z 420i........................................................... .....................................120
A .5 R IE G L L M S-Z 390 ........................................................................................ 12 1
A .6 R IE G L L M S-Z 2 10ii .............. ................................................... ......... .... ..... 12 1
A .7 T rim ble G S 10 1 ....................................................... ........................ 122
A .8 Trim ble GX 3D ................................................... ................ ......... 122
A .9 M inolta V IV ID 910 .............................................. .. .. .... .. ....... ....... 123









A.10 Zoller-Frohlish IM AGER 5006 ...........................................................................123
A .11 IQ Sun 880 ................... ........ ........ ..... ........................... 124
A.12 Comparison of Terrestrial LASER Scanner Specifications............................124

L IST O F R E F E R E N C E S ............................ ...................................................... .....................126

B IO G R A PH IC A L SK E T C H ............................ .................................................. ................... 132
















































8









LIST OF TABLES


Table page

1-1. Phase difference ranging principles expressions and equations. .....................................31

1-2. Time of flight ranging principles expressions and equations. .............. ......... ..........33

1-3. Optical triangulation ranging principles expressions and equations...............................34

1-4 L A SE R classifi cation .............................................................. .....................................36

5-1. Control points used for the geo-referencing of the March 23, 2006 dataset....................77

5-2. Point clouds coregistration RM S values. ........................................ ........................ 77

5-3. Summary of volume change computations......................... .......................... 80

6-1. Soil roughness collected datasets.......................................................................... ...... 86

6-2. Definition of soil roughness parameters. ........................................ ....................... 89

6-3. Soil roughness parameters results from random profiles for the Citra 02 grid..................91

6-4. Soil roughness parameters results from random profiles for the Hastings grid ...............91

6-5. Soil roughness parameters from 3D surface models............................... ...............95

6-6. Comparison of soil roughness parameters for Citra 02 from 3D surface models and
random generated profiles.............................................. .................... ............... 95

6-7. Comparison of soil roughness parameters for Hastings 02 from 3D surface models
and random generated profiles.................................................. ............................. 95

6-8. Comparison of soil roughness metrics obtained from the traditional and alternative
m ethod.............. ........................ .................................... .................. 97

7-1. M -TLS data set geo-referencing control network..........................................................104

7-2. Geo-referencing residuals analysis. ........................................................................... 105

7-3. Diameter and heights measurements for stem volume estimation................................113

A-1. Comparison of terrestrial laser scanner specifications ....................................................124









LIST OF FIGURES

Figure page

1-1. T riangulation netw ork ......... ................. ............................................................... 17

1-2. E early E D M equipm ent. ........................................................................... .....................20

1-3. K+E RangeMaster III EDM unit ......... ....... ............................ ...............24

1-4. Spectra-Physics G eodolite. ...................................................................... ....................25

1-5. NASA AOL system. ........................................ ....................27

1-6. Phase difference ranging principle.......................................................... ............... 31

1-7. Tim e of flight ranging principle...................................................................... 32

1-8. Optical triangulation ranging principle .................................... ............................ ........ 33

1-9. Scanning axes on a panoramic view scanner .......................................... ............... 34

2-1. Survey vehicle with a similar concept to MOBLASS................................................. 41

2-2. O PTECH ILRIS block diagram .............................................................. .....................42

2-3. Mobile Terrestrial Laser Scanning (M-TLS) system: truck, lift and ILRIS....................44

2-4. M-TLS deployed in Georgia, showing the power generator and field computer on the
tru ck b e d ................... ...................4...................5..........

2-5. M -TL S pan tilt bases........... ..................... ... ........ ...................... .. .. .... .... 46

2-6. The ILRIS unit with on axis video camera and GPS antenna.........................................48

2-7. Array of geodetic quality GPS base stations used for geo-referencing. ..........................48

2-8. Installation of the tiltmeter unit on the instrument frame. ........................................49

3-1. Screen capture of the ILRIS unit controller software during scanning operations............51

3-2. Different configurations of the M-TLS system during several data collection
proj ects ............ ..... .............................................. ............................ 5 1

3-3. Screen capture of the Parser software during the setup of parsing settings..................52

3-4. V visualization of LID AR data ....................... ....... ................................... ............... 54

3-5. Prim itive fitting process illustrated ..... .. ............................................................ 55



10









3-6. Cross section generation from the point cloud. ..................................... ............... 55

3-7. Point cloud m erging exam ple. ................................................ ............................... 57

3-8. Gridding operations, from point cloud to grid. ........................................ ............... 59

3-9. Examples of advanced mathematical operations in the processing of Airborne
L ID A R data ....................... ......... .................. ................................ 60

4-1. Traditional and alternative methods for measuring and recording spatial information
in archeological and paleontological sites. ............................................. ............... 63

4-2. Rendering of the dig site point cloud macro model showing laser return intensity. .........64

4-3. RGB textured renderings of high resolution point clouds. ...........................................64

4-4. 3D surface grids used to compute the volume of dirt extracted in one day....................65

4-5. Traditional and alternative ways to perform geological field mapping...........................66

4-6. Geo-referenced point cloud rendering of the Tuolumne quarry..................................... 68

4-7. Shaded relief image from a gridded model of the south wall of the quarry. ..................68

4-8. Methods of measuring alligators and crocodiles. .......... ........ .................... 69

4-9. M ethods for generating beach profiles. ........................................ ......................... 70

4-10. M ethods for deriving soil roughness metrics................. ....... ...... ................... .............. 71

4-11. M ethods for estimating forestry metrics. .......................................................................... 72

5-1. Beach erosion hot spot study site location .............................................. ............... 76

5-2. Rendering of the M arch 23rd, 2006 dataset.. ......................................... .... ...............77

5-3. Features used to check the co registration of the point clouds. ...................................78

5-4. Image maps from the 10 cm elevation grids ........................................... ............... 78

5-5. Image maps from the elevation change grids. ...................................... ............... 79

5-6. Beach line and crest of berm position plots for each of the data collection dates ...........81

5-7. Beach profiles extracted from the grids showing the recession of the berm .................82

5-8. Comparison of profile resolution generated from traditional methods and M-TLS .........83

6-1. D ataset reprocessing steps.. .............................. ... ......................................... 87









6-2. Renderings of the 1 cm elevation grids........................................ .......................... 88

6-4. Roughness parameter plots for the Citra 02 dataset parallel to the X axis ......................90

6-5. Roughness parameter plots for the Citra 02 dataset parallel to the Y axis ......................90

6-6. Roughness parameter plots for the Hastings dataset parallel to the X axis .....................90

6-7. Roughness parameter plots for the Hastings dataset parallel to the Y axis .....................91

6-8. Citra norm alized height autocorrelation. ........................................ ....................... 92

6-9. H astings height autocorrelation. ............................................................. .....................93

6-10. Correlation lengths extraction for the Citra 02 dataset. .............................................. 93

6-11. Correlation lengths extraction for the Hastings dataset. .................................................94

6-12. Comparison of experimental correlation length distributions with respect to the
assume ed norm al distribution. ...................................................................... ..................96

6-13. M eshboard used to digitize the soil surface transect. .............................. ......... ...... .96

6-14. Plots of meshboard derived data................................ ........................ 97

7-1. A erial photographs of the test site. ........... ............... .......... ........ ...............103

7-2. Shaded relief digital elevation model rendered from the airborne laser scanner data
of the test site. .......... ... ............... ................................... ..........................104

7-3. Rendering of the fused point cloud, color coded by elevation. ....................................106

7-4. Rendering of fused point cloud cross section in the along the flightline direction.......... 106

7-5. Rendering of fused point cloud cross section in the cross flightline direction................106

7-6. Rendering of the fused point cloud, grey scale from the laser return intensity ..............107

7-7. Rendering of the fused point cloud, color coded by elevation + laser return intensity. ..107

7-8. Rendering of the top view of fused point cloud, color coded by elevation + laser
retu rn in ten sity ........................................................................ 10 8

7-9. Rendering of the "Forest Cube"........................... .. ....... ................................. 109

7-10. Rendering of point cloud used for stem counts .................. ....................................109

7-11. Fitting of a circle at breast height for determining DBH and stem location..................10









7-12. Single tree height determ nation. ........... ............... .. ....... .................................11

7-13. Diameters at different heights for volume computations.......................................... 112

7-14. Individual tree m etric m easurem ent......... ............................................. ............... 114

A -1. O ptech IL R IS T O F T L S ......................................................................... ................... 119

A -2. Leica H D S 3000 TO F TL S ......... .................. .............. ..................................... 119

A .-3. L eica H D S 4500 PD TL S ....................................................................... ...................120

A -4. R iegl LM S-Z 420i TO F TL S ............................................................................ ............120

A-5. Riegl LM S-Z390 TOF TLS ...................................................................... ............... 121

A -6. Riegl LM S-Z210ii TOF TLS ............................................... ......... .. .................. 121

A -7. T rim ble G S 10 1 T O F T L S ........................................................................ .................. 122

A -8. T rim ble G X 3D T O F TL S ....................................................................... ..................122

A -9. M inolta V IV ID 910 O T TL S ......... ................. ................................ ...........................123

A-10. Zoller-Frohlish IM AGER 5006 PD TLS ........................................ ...... ............... 123

A-11. IQSun 880 PD TLS.............. .... ................................. .. 124









Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

SCIENTIFIC APPLICATIONS OF THE MOBILE TERRESTRIAL
LASER SCANNER (M-TLS) SYSTEM

By

Juan Carlos Fernandez Diaz

August 2007

Chair: Ramesh Shrestha
Cochair: K. Clint Slatton
Major: Civil Engineering

Terrestrial Laser Scanners (TLS) are mapping instruments composed of a Laser

rangefinder and an optical-mechanical system used to steer the laser beam across the surface of

interest. The University of Florida Geosensing Engineering and Mapping (GEM) Research

Center is working towards developing a Mobile Terrestrial Laser Scanner (M-TLS) system. The

core of the M-TLS is a commercial 2-axis ground based laser scanner which is integrated to a

mobile telescoping, rotating, and tilting platform which provides up to 6 degrees of freedom for

performing scanning operations. A scanner built-in 6 megapixel digital camera and a digital

video camera provide the M-TLS moving and still imaging capability. At the final stage of

development the M-TLS positioning and navigation system will include a differential GPS array,

tilt sensors and an inertial measuring unit which will allow data collection and georeferencing in

both static and dynamic modes.

The M-TLS laser scanner is capable of generating 3D spatial and multispectral data sets. A

typical dataset is composed of a cloud of millions of points for which 3D coordinates, laser

intensity and/or RGB information are available for each one. Data can be collected in a range

from 3m to 1500m for a target with an 80% reflectivity or 3m to 350m to targets with a 4%









reflectivity. The laser operates at a wavelength of 1535 nm, with a pulse width less than 10 ns

and energy of less than 10 [tjoules. The sampling separation can be adjusted down to 0.00115,

and the scanning speed is 2,000 points per second.

The M-TLS is a unique tool that enables GEM researchers to acquire high density point

clouds from an advantageous terrestrial geometry, being a very valuable complement for

Airborne Laser Scanner data sets. The applications of the M-TLS data sets are numerous in both

the fields of science and engineering. Tested applications by the GEM center include urban

mapping, as-built surveying, building damage assessment, bridge load analysis, forestry metrics

extraction, beach erosion mapping, paleontology and archeology dig mapping, structural geology

mapping, forest fire fuel estimation, soil spatial characterization and vehicle 3D modeling. This

thesis centers around the novel applications of the M-TLS to specific scientific problems:

thoroughly analyzing the applications to beach erosion hot spot mapping, soil roughness metrics

extraction, and forestry metrics extraction.









CHAPTER 1
INTRODUCTION TO TERRESTRIAL LASER SCANNING

1.1 Geodesy and the Need for Measurements

Since early history mankind has had questions about nature, many of which can be

answered if the proper measurements are made. The shape and size of the Earth was one of the

first questions that humans asked. The earliest answers to those questions were based on folklore,

common sense or primitive scientific method. These earliest conceptions had to do with

supernatural creatures or deities and varied from culture to culture. The Greek philosophers did a

lot of thinking about this issue. Homer, the epic Greek writer, popularized the conception that the

Earth was a flat disk. It was around the time of Pythagoras, the Greek Mathematician who lived

between 580 and 500 BC, that the idea that the Earth must be a sphere, the most perfect three

dimensional geometric figure, took form. Following this Pythagorian principle other

philosophers such as Aristotle and Archimedes tried to estimate the size of the Earth (Burkard,

1983; Smith, 1996).

The first scientific approach to measure the size of the Earth is credited to Eratosthenes

(Burkard, 1983; Ewing & Mitchell, 1970; Smith, 1996), a Greek philosopher who also was a

librarian in Alexandria. By studying the difference in the lengths of shadows cast by the Sun on

the same day in the cities of Alexandria and Syene, and assuming that the Earth was a sphere, he

was able to determine the angular separation between these two cities. Then by measuring the

linear distance along the surface between the two cities he was able to estimate the size of the

spherical Earth. The thoughts and works of these Greek philosophers led to the birth of the

discipline that is called geodesy. Modern geodesy is defined as a branch of applied science that

deals with the determination of the size, shape and orientation of the Earth and its gravitational

field, and the variations of these parameters with time (Ewing & Mitchell, 1970).









A central focus of geodesy is determining the locations of points on the Earth, i.e.,

determining the coordinates of points relative to a well defined origin and reference frame. The

evolution of surveying techniques and instrumentation led geodesists to treat positioning as two

separate but related problems (Shrestha, 1983). For determining the horizontal position

components, astronomical observations, triangulation, trilateration and traversing were some of

the techniques employed. For determining the vertical position geodetic leveling, trigonometric

leveling, barometric leveling or echo sounding techniques are employed (Smith, 1996).

As described in this introduction, the foundations of geodesy are math and physics,

however, these rely on real measurements to model the world. The most basic geodetic

measurements are the ones pertaining to time, distances and angles. If close attention is given to

the history of geodesy it becomes clear that great leaps in knowledge came when new technology

and its attendant instrumentation permitted leaps in the accuracy and precision of the

measurements. Measuring the distance between widely separated points on Earth has always

presented a challenge. One of the earliest successful methods of connecting widely separated

points is a method known as triangulation. In a triangulation network, the angles of a chain of

triangles are observed, along with one relatively short "baseline" length. From this single

distance observation, and using trigonometry, all the sides of the triangles are computed.


C










Figure 1-1. Triangulation network. The baselines A to B, as well as all the internal angles need to
be measured to determine the A-C distance.









Triangulation is a time consuming method and errors in the angular and distance

measurements propagate all along the network, reducing the precision and accuracy of the

desired A to B distance. A breakthrough in distance measurements came in 1941, when a

Swedish geodesist named Eric Bergstrand conceived a new technique to measure the time it took

a beam of light to travel a known distance to determine the speed of light. He then realized that if

the speed of light were accurately known, he could invert that technique and measure the

distance between two points (Carter, 1973). This was the dawn of Electronic Distance

Measurement or EDM. Because of the curvature of the Earth, it still remained necessary to

determine the distance between widely separated points with a series of shorter distance

measurements, in a method known as trilateration. The ease of use, accuracy and productivity of

EDM instruments allowed trilateration to quickly displace triangulation.

Since 1941 EDM has evolved from a concept to a proven technology. Today there are

many different forms of EDM instruments; some use radio frequencies, while others use light

waves. EDM is used to measure distances small and astronomical, from micro structures to the

distance between the Earth and the Moon, and even to neighboring planets. EDM has fulfilled

the need for accurate distance measurements and has provided scientists and engineers with the

data they need to build a model of our world.

During the past decades EDM instruments have been developed that include opto-

mechanical scanners which steer the measuring beam over a selected pattern to collect a set of

surface coordinates that can be used to create a mathematical representation of any surface in

three-dimensional space. This thesis will present the experiences and results of almost two years

of experimenting with a terrestrial LASER scanner to fulfill the need for measurements in

several scientific fields such as forestry, soil science, geology and beach morphology.









1.2 History of LASER EDM and Scanners

In tracing the origins of LASER EDMs and scanners one can dig as deep as to the early

astronomical observations aimed to estimate the velocity of light, performed by Romer and

Huygen, or the ground-based experiments conducted by Hippolyte Fizeau, Leon Foucault, Simon

Newcomb and Albert A. Michelson. However, most historians will set the origins of the EDM

technique and instruments around the 1940s, highly influenced by the wartime efforts to develop

the RADAR (Radio Detection and Ranging).

The origin of the first EDM instrument began in 1938 when the physicist and geodesist

Erik Bergstrand, of the Swedish Geographical Survey Office, began to investigate the

possibilities of using a Kerr cell as an electro-optical shutter to modulate a beam of light in an

attempt to measure of the speed of light. Bergstrand's first operational instrument was reported

to work in 1941. It used a Kerr Cell controlled by a crystal oscillator to modulate light from an

ordinary incandescent light bulb. The light beam was collimated and projected by a parabolic

mirror to a reflective corer cube array, the returning waveform was detected by a

photomultiplier tube (PMT), and the round trip travel time was determined from the difference in

phase of the transmitted and reflected modulated light beam (Carter, 1973; Smithsonian, N.D.).

In 1947 Bergstrand took his instrument to a 6 km baseline in Orland and obtained a

measurement of the speed of light of 299,793.1 +0.2 km per second. A year later in August 1948,

Bergstand read a paper at the meeting of the International Association of Geodesy(IAG) held in

Oslo, Norway. In that paper he explained that one could reverse the process, measure the light

time of flight and use the value of the speed of light to compute the distances between two

points. Soon after the meeting, Bergstand licensed the concept to the Swedish AGA (Svenska

Aktiebolaget Gasacumulator) company to develop a commercial instrument.









AGA produced the first EDM instrument in the early 1950s, and marketed it as the

"Geodimeter", short for geodeticc distance meter". The instrument used a Kerr cell to modulate

the light, but the incandescent light bulb used by Bergstand was replaced with a mercury vapor

light. The development of the Geodimeter by AGA continued through the 1950s and 1960s. The

last model to use mercury lamps was the Geodimeter Model 6 introduced in 1964. Figure 1-2

shows an early production Geodimeter and its required corer cube reflector array. (NOAA,

N.D.; Smithsonian, N.D.)

A B











Figure 1-2. Early EDM equipment. A) AGA Geodimeter B) Corer cube reflectors. (Source:
http://pubs.usgs.gov/gip/monitor/techniques.html Last accessed March 16th, 2007)

Around the same time that AGA was producing the first Geodimeter, Harry A. Baumann

of the South African Trigonometrical Survey and Trevor Lloyd Wadley of the

Telecommunications Research Laboratory of the South African Council for Scientific and

Industrial Research (CSIR) were developing the Tellurometer. The Tellurometer was the first

commercial EDM instrument to use microwave beams to measure long distances with geodetic

accuracy. The first model appeared in 1954, marketed as the Micro-Distancer M/RA 1. It was

composed of two units, designated as master and remote, each set on the extreme points of the

distance to be measured. Its range was between 30 and 50 km. The system used a continuous

3 GHz carrier frequency modulated by 10 megahertz and three other nearby frequencies. The









remote station retransmitted the incoming wave in a similar wave of more complex modulation,

and the resulting phase shifts of the carrier and the modulating signals was used to

unambiguously determine the distance traveled(NOAA, N.D.; Smithsonian, N.D.). The

Tellurometer had the disadvantage that propagation of the microwave energy caused multipath

reflections that degraded the system precision and accuracy.

Between 1938 to 1960 EDM evolved from a concept to widely used operational technique.

However, its greatest leap in range and accuracy was yet to be realized. In 1954, Charles Townes

at the University of Columbia invented the MASER (Microwave Amplification by Stimulated

Emission of Radiation). (IEEE, N.D.) A maser is a cavity filled with gas (the first used ammonia

gas) that when "pumped" with microwave radiation generates more microwave radiation.

1957 will always be remembered as the year that the Soviet Union launched the first

artificial satellite Sputnik I. But in November of 1957 Gordon Gould, a graduate student at

Columbia University working with Townes, coined the acronym LASER, for Light

Amplification by Stimulated Emission of Radiation, and described the principal components of

the LASER. However, Gould did not publish his work, focusing his efforts in finding a position

and the resources to try to build the LASER (IEEE, N.D.; Taylor, 2000).

In March 1958, Arthur Schawlow, apparently independently, also realized that the secret to

the LASER involved an optical cavity, along the lines of a Fabry-Perot interferometer.

Schawlow shared his idea with Charles Townes, his brother-in-law, and together they wrote a

paper entitled "Infrared and Optical Masers" published in Physical Review Volume 112, number

6. (Schawlow & Townes, 1958)

The paper by Schawlow and Townes encouraged widespread thinking about how a Maser

at optical wavelengths or LASER might be built. In 1960 Theodore Maiman and his colleagues









at Hughes Aircraft Company, succeeded in building the first solid state pulsed LASER, using a

Ruby Rod. The LASER light was red with a wavelength of 0.6943 micrometers. That same year,

Ali Javan and his colleagues from Bell Laboratories succeeded in building the first gas LASER.

The HeNe LASER produced a continuous beam at five different wavelengths, achieving the

highest power of 15 miliwatts at 1.153 micrometers (Javan et al., 1961 & Bennett, 2000).

The gas LASER was well suited for use in terrestrial EDM instruments. The light produced

by a LASER is highly mono-chromatic and coherent (the photons of a LASER beam have a

single wavelength, phase and move in the same direction). These attributes allow a LASER beam

to have a small divergence, which means that the energy does not spread in the typical large

spherical pattern of other light sources. Replacing the mercury vapor light with a HeNe LASER

dramatically increased the operating range of the Geodimeter, and enabled the development of

other smaller EDM instruments that quickly took over many aspects of surveying. The solid state

pulsed ruby LASERs were not very energy efficient, and even the best "Q-switched" devices

produced light pulses meters in length, making them poorly suited for short range geodetic

surveying, but they could produce large quantities of energy, however, they would soon find a

role in space geodesy (Carter, 1973).

For geodesists the main limitation of the improved LASER based EDM instruments was

not the technical capabilities of the instruments themselves, but rather that the curvature and

topography of the Earth limited the line of sight distances. As soon as the first artificial satellite

was put in orbit in 1957, geodesist start to realize that a spaceborne "target" could greatly extend

the measured baseline distances. At the same time scientists interested in the fields of gravity and

relativity (which also are of great importance to geodesy) started working on a concept to

employ a high density and high altitude artificial satellite to measure slow changes in the









universal gravitation constant (G) by accurate tracking the satellite path using retroreflectors and

pulsed search lights (Bender et al; 1973).

In 1964 the first geodetic satellite (Beacon Explorer 22-B) was put in orbit, it had an array

of corner cube reflectors that were illuminated using pulsed ruby LASER beams, the first

ranging measurements obtained on Oct 31, 1964 (Carter, 1973; McGarry & Zagwodzki, 2005).

However, even before the satellite was launched scientist realized that low orbiting satellites

imposed several challenges such as very short visibility times and Earth's gravitational

perturbations that would limit the quality of the relativist experiments.

As early as 1962 J.E. Faller had proposed the idea of placing a retroreflector on the surface

of the Moon, which could be used to bounce back a LASER beam shot from the Earth. Between

1962 and 1964 experiments that included the detection of LASER beams bounced from the

moon's surface were performed, and in 1965 the Lunar Ranging Experiment (LURE) multi-

institutional team was formed. From 1965 to 1969 the LURE team focused their efforts to

develop the largest and most sophisticated EDM system to date. Their first great milestone was

reached on July 21st, 1969, when Neil Armstrong aligned and leveled the first corner cube

reflector array on the surface of the moon.

Shortly after the installation of the array on the surface of the Moon, scientists on the Earth

used the Lick Observatory's 3.05 meter telescope and a Q switched pulsed ruby LASER to aim a

2 arc minute divergence beam to the array. The first successful return signal from the array was

obtained on August 1, 1969 at Lick Observatory, shortly after, on August 20th the McDonald

Observatory reported success obtaining returns. Successful results were also reported that same

year by the Air Force Cambridge Research Laboratories (AFCRL) Lunar Ranging Observatory

in Arizona (Bender et al., 1973; Carter, 1973).









Down on Earth, AGA continued its development of the Geodimeter, introducing its Model

8 in 1967, which was its first instrument to use a helium-neon LASER. The LASER allowed the

extension of the measuring range of the lamp units of 20 to 30 km to a range of 60 km in both

day and night conditions. (Smithsonian, N.D. & Cheves, M; 1999)

In 1965 LASER Systems & Electronics, Inc. was established by a team of physicists and

engineers who had worked at the Engineering Development Center at Arnold Air Force Base in

Tullahoma, Tennessee. In 1970, LASER Systems & Electronics unveiled their first electronic

distance measuring instrument: the Ranger. This was the first competition AGA faced; it used a

red visible LASER and was capable of ranging distances from 1 meter to 6 km with an accuracy

of 5 mm +2 ppm. The EDM side of LASER Systems & Electronics was sold to Keuffel &

Esser in 1971, which continued to manufacture the Ranger, RangeMaster and AutoRanger

(1977) series. (NOAA, N.D.; Smithsonian, N.D.)












Figure 1-3. K+E RangeMaster III EDM unit. (Source:
http://celebrating200years.noaa.gov/distancetools/ranger.html, last accessed March
20th, 2007)

During the 1970's several surveying and electronic instrument companies developed EDM

equipment. Among those companies were Cubic Corporation, Hewlett-Packard, Wild and Zeiss.

One of the trends of this period was to combine an angular measuring device with an EDM into

what was, and still is called a total station. These instruments continued to evolve into the

simpler, more compact, accurate and cheaper units that can be found today. Most of these units









require the use of corner cube reflectors to get a strong return signal that enable a computation of

the distance. Some EDM devices followed a different evolutionary path, and they are able to use

weak return signals that are bounced back from natural targets such as the surface of the Earth or

man made structures to compute the range from the instrument to the surface.

Between 1964 and 1966 Spectra-Physics, a company based in Mountain View, California

developed a series of precise LASER-based EDMs. Its first prototype model, the Mark I, was

designed as an airborne profile recorder (1964). The Mark II model was mounted in a Douglas

A-26 and used to record a height profile across a stadium from a flying altitude of 300 meters.

The Mark III model introduced in 1966 was marketed as a "Geodolite", its development was

funded by the United States Army Engineer Topographic Laboratories. (Smithsonian, N.D.).

This concept was further used not only as a profiler but also as a LASER altimeter and for

bathymetric measurements.














Figure 1-4. Spectra-Physics Geodolite. (Source: http://historywired.si.edu/object.cfm?ID=22, last
accessed March 20th, 2007)

In 1969 Hickman & Hogg were the first to report results on the feasibility of using a pulsed

blue-green (frequency doubled NdYAG: Neodynium Yttrium Aluminum Garnet) LASER from

an airborne platform for near-shore beach reconnaissance surveys over the shores of Lake

Ontario. The concept was employed on the LASER altimeter experiment which was flown on the









Apollo 15, 16 & 17 (1971, 1972 & 1972) to the Moon. The LASER altimeter experiment

obtained data on the altitude of the Command Service Module (CSM) above the lunar surface.

The altimeter was used to support mapping and panoramic camera photography. It operated in

two modes. In the independent mode it performed ranging measurements every 20 sec. In the

slave mode when the metric camera was operated, it automatically emitted a LASER pulse to

correspond to a midframe range for each frame (NASA, N.D.).

Also between 1971 and 1972 U.S. Naval Oceanographic Office (NAVOCEANO)

performed flight tests of a prototype airborne LASER bathymetry profiler known as Pulsed Light

Airborne Depth Sounder (PLADS). Even though those tests proved that airborne LASER

bathymetry was feasible they were not conclusive in terms of operational system performance

and cost benefit ratio. In 1973 a joint project between NASA and the Naval Oceanographic

Office was established to develop and thoroughly test an Airborne LASER Bathymeter (ALB)

System. The construction of the system was concluded in 1975. Using two pulsed LASER

sources: a 2 kW 540 nm 6 nsec 10-200 pps Neon Ion LASER and a 2MW 532 nm 8 nsec 10-50

pps frequency doubled Nd Yag LASER. The receiver consisted of a 28 cm Cassegrain telescope,

two narrowband 0.4 nm filters and an 8575 RCA photomultiplier tube. Over 200 hours of fight

test were used to collect bathymetric data on Chincoteaque, VA, the Chesapeake Bay and Key

West, FL. (Kim, Cervenka, Lankford; 1975)

In 1975 NASA Wallops Flight Center and AVCO Everett Research Laboratory proposed a

bathymetric LIDAR Airborne system building on the previous experiences of the ALB. The

proposal passed in 1977, and the new instrument was built in the same year by the AVCO

Everette Corporation. The system became known as Airborne Oceanographic LIDAR (AOL).

The program was sponsored by NASA, NAVOCEANO, and NOAA as an advanced testbed









sensor for altimetry, bathymetry, hydrography, and fluorosensing research. (Guenther et al.,

1978; Mitchell, N.D.).

The first implementation of AOL was based on AVCO C-5000 gas (neon/nitrogen)

540.1 nm 2 kW 400 pps LASER, a 56 cm nutating scanner mirror, a 30.5 cm diameter

Cassegranian f/4 telescope, a narrow band 0.4nm filter and a photomultiplier tube. This was the

first implementation of a scanning airborne LIDAR altimeter. (Guenther et. al., 1978) The AOL

was improved and redesigned several times over two decades. Improvements in enabling

technologies such as differential GPS, Inertial Measuring Units (IMU) and best estimate

trajectory algorithms continually enhanced the precision, accuracy and productivity of the AOL

system. In 1994 the AOL underwent major hardware renovation and its topographic mapping

and ocean fluorosensing functionalities were separated in to two new systems: The Airborne

Topographic Mapper (ATM) and the AOL.

A i B







Figure 1-5. NASA AOL system. A) P3 Orion platform. B) Scannig mirror. C) Instrument rack.
(Source http://sealevel2.jpl.nasa.gov/jr_oceanographer/oceanographer-williams.html,
last accessed March 20, 2007)

The key technological advances that enabled the simple LASER profilers to become

scanners were the development of more sensitive photodetectectors in the form of

photomulipliers and avalanche photo diodes (which enabled the recovery of the reflected LASER

pulses at lower signal-to-noise-ratios), LASER beam steering mechanisms (scanners and their

respective encoders), and finally the the highly accurate Integrated Navigation Systems (INS).









The greatest advantage of airborne LASER scanning over original profiling methods is the

dramatic increase in mapping coverage per unit of flight time.

Before 1995 all airborne LIDAR systems were custom built, highly expensive, and were

only available for big research institutions and companies. However in 1995 Optech Inc., a

company based in Toronto, Canada, offered the first commercial-off-the-shelf Airborne LIDAR

Mapping System. Soon after other companies started offering their own systems: Riegl in 1996

with its low altitude scanner, Saab Survey Systems in 1997, and the Top Eye and Azimuth

Corporation in 1998 with its Aeroscan system (which later was bought by Leica)

(Baltsavias, 1999; Flood,2001; Cheves, 2002).

As technology advanced the receiving electronics increased in sensitivity and timing

accuracy, LASERs became smaller yet more powerful and robust, scanners became smaller and

their encoders more precise. The continued miniaturization and improvements of the airborne

LASER scanners led the manufacturers to consider the construction of small, short range LASER

scanners that could be mounted on tripods or small vehicles. The first units were introduced in

the late 1990's. Some referred to these units as terrestrial LASER scanners to differentiate them

from their airborne counterparts. Today there is no universally accepted nomenclature as each

manufacturer has created its own terminology. Riegl names them "3D Imaging Sensor", Optech

calls them "LASER Ranging and Imaging Systems", Leica labels them "High-Definition

Surveying Systems", but in the government and academic literature they are mainly referred as

terrestrial LIDAR scanners.

Over the last 8 years Terrestrial LASER Scanners (TLS) have become very common and

useful tools, mainly used in the areas of architecture, civil engineering, surveying and mapping.

As mentioned earlier, science is always in need of measurements to model the world. Just as









their predecessor, the Airborne LASER Scanner, proved to be useful to different branches of

natural science, TLS systems have great potential to provide dense and accurate range sampling

of the environment in an efficient fashion.

1.3 TLS Subsystems and Principles of Operation

Terrestrial LASER Scanners are complex and very precise instruments; they are composed

of two basic subsystems. The first subsystem is a LASER ranging device commonly called

LIDAR for (Light Detection And Ranging) or LADAR (LASER Detection And Ranging). The

latter is the nomenclature generally used by the military. The second subsystem is an optical and

or mechanical device capable of steering the LASER beam in a scanning fashion over the area of

interest. (Frohlich & Mettenleiter, 2004) Even though TLS have these common elements, there

are at least a dozen different types of instruments in the current market. This is due to the fact

that there are 3 different ways a LIDAR can work and more than 3 ways to do scanning in two

dimensions.

1.3.1 LIDAR Ranging Principles

A LIDAR can be used to measure the distance between two points in any of the three

following ways: phase difference, time-of-flight and by optical triangulation. Each of these

ranging approaches has its own set of strengths and weaknesses. In the phase difference units,

the continuous LASER signal can be modulated at very high frequencies, and the numbers of

ranges per second is generally limited only by the speed at which the data can be recorded. Thus

PD units have the largest point collection throughput of all types of units. Their main limitation

is the ambiguity of the range measurement. PD units are ideal in situations where very short

acquisition times are desired, with sample collection rates in the order of hundreds of thousands

of samples per second and mm level accuracy, but with limited range.









Optical Triangulation units are ideal for measuring distances of a few meters with

micrometer level accuracy at high data rates. However, its accuracy depends on the relation

between range and baseline distance and falls off rapidly with increasing range. Other limitations

are that its performance can be degraded if the surface is not uniform in shape or reflectance and

by the presence of noise in the form of exterior illumination from non-coherent light

sources.(Curless & Levoy;1995)

Time of Flight units have the advantage that they provide unambiguous ranges from a few

meters all the way to thousands of meters. However, because after they emit the LASER pulse

they must wait until there is a return signal before they can send the next pulse. The point

collection throughputs are relatively low compared to the PD and OT units. However, is worth

mentioning that there are some special airborne and long range LIDAR systems that are able to

work with multiple pulses at a given time. The other limitation is that the range resolution (the

ability to separate proximate object at different ranges) decreases as the pulse width increases.

That means that for pulsed LIDAR there is always a tradeoff between range and range resolution.

However they are established as the most common type of ranging LIDARs.

1.3.1.1 Phase difference measurement (PD)

Phase Difference was the ranging method used in early geodetic EDMs like the

Geodimeter & Tellurometer and on current systems that employ continuous wave (CW)

LASERs. In these LIDARs the amplitude of the LASER "wave" is modulated and the phase

difference between the outgoing and reflected wave is measured. The problem with this method

is that phase differences are not unique, there is always an ambiguity about the number of

complete modulating wave cycles that have occurred prior to the phase difference. Most current

TLS that work under this modality do not provide ambiguity resolution so they have a limited

range usually less than 100 meters (Wehr & Lohr, 1999 ;Frohlich & Mettenleiter, 2004). It is









important to clarify that the LASER has its own natural frequency and wavelength. Optical

LASERs have wavelengths in the range of .4 to 1.5 |tm. The phase difference that is used to

determine the target range is based on the amplitude modulating signal (not the LASER natural

wavelength), which has a wavelength in the order of several to hundreds of meters.

Outgoing Waveform


TLS


Time


r-r .I ,




J-


I Phase
Difference
Figure 1-6. Phase difference ranging principle.

Table 1-1. Phase difference ranging principles expressions and equations.


Parameter


Formula


Ambiguous phase difference
Unambiguous phase difference
before one full wave cycle
2 way travel time before one full
wave cycle
Long amplitude modulating signal
wavelength
Short amplitude modulating signal
wavelength
Speed of light
Signal to noise ratio
Maximum unambiguous range

Range

Range resolution

Single shot range accuracy


APD = 2r x n+



0 f Along
2way X f2camer 2X c

Along
J earner

Short
c
s/n

RD Along
Rmax 2
1 l r ong
R=-xcx --xfe --xi
R -XCX carrier -
2 2I 4/r
AR = hoxA0
47r
4sIho,, 1
OCR X
44r x -n


A A I I A I I I I
AAAWMH 1111M
V


11FIf IHI VVWMYOM
A AA h I j A T, I

V









1.3.1.2 Time of flight (TOF)

In PD LIDARs the phase difference of the continuous LASER is used to determine the 2

way time-of-flight of the modulated signal. As mentioned in the previous section, the main

disadvantage of this method is the ambiguous range. An alternative that help overcome this

limitation came with the development of Q-switching by McClung and Hellwarth in 1961

(McClung & Hellwarth, 1962). This invention enabled the emission of very energetic LASER

pulses rather than the continuous wave beams. This allowed to directly measure the 2 way time-

of-flight without any ambiguity. However, even when these pulses are relatively short in time,

generally in the order of a few nanoseconds, at the high speed that light travels this translates into

several centimeters (e.g. 10ns = 3.0 m). In order to obtain the sub-centimeter accuracy an

specialized electronic circuit called a Constant Fraction Discriminator (CFD) is used to precisely

time a specific point on the pulse (generally the half point of the pulse amplitude). With discrete

packets of light and the CFD is a simple matter to measure with high accuracy the time

difference between the emission of the pulse and the detection of its reflected return. This

method provides unambiguous range measurements of distances limited only by the dispersion

of the LASER energy and the sensitivity of the detector (Wehr & Lohr, 1999).


Outgoing Waveform




Inco 0 g Waveform

TLS ____

Time Difference Time
Figure 1-7. Time of flight ranging principle.









Table 1-2. Time of flight ranging principles expressions and equations.
Parameter Formula
2 way travel time t
Pulse width At
Speed of light c
Signal to noise ratio s/n
Pulse rise time tpu en
Range R 1
R=-xcxt
2
Range resolution 1 A
AR = x c x At
2
Single shot range accuracy c x tpulse nse 1
O"R ~ -- --- X I --
2 ,-vs/n

1.3.1.3 Optical triangulation (OT)

In PD and TOF the outgoing and incoming LASER beam follow the same optical path,

however on OT the reflected waveform is observed from a different vantage point. In OT units a

LASER beam is steered by a scanning mirror over the target surface and its reflection is

collected through a lens that focuses an image on a position sensitive detector such a CCD array.

The position of the spot image on the pixels of the camera, the scanner angle and the LASER to

lens optic center baseline is then processed using trigonometry to determine the distance to the

target. (Beraldin et al., 2003)


Figure 1-8. Optical triangulation ranging principle.


Detector









Table 1-3. Optical triangulation ranging principles expressions and equations.
Parameter Formula
Baseline distance D
Deflection angle of the LASER beam a
Position of the imaged spot p
Focal length f
Range Dxf
p + f x tan(a)
"Horizontal" position X = Z x tan(a)
Position accuracy Op
Single shot ranging accuracy Z2
fxD p

1.3.2 Scanning Mechanisms

Traditionally, geodetic scanning instruments have used reflective optics coupled to a

mechanical system, although some newer instruments (CATS & Jigsaw) use refractive scanning

elements such as Risley prisms (Carter et al., 2005; Marino & Davis, 2005). Current TLS

systems have a 2 axis capability. This is achieved in its simplest implementation by a single line

scanner through a rotating or nutating mirror, and the second scanning axis is obtained by

rotating the complete instrument as shown in Figure 1-9.

Azimuth Axis

C7)











Rotating Mirrr
Figure 1-9. Scanning axes on a panoramic view scanner. (Adapted from original source at:
http://www.riegl.com/terrestrial_scanners/lms-z390_/390_all.htm, last accessed
March 20, 2007)









This type of scanner is usually called a Panoramic View Scanner and is capable of

scanning 3600 in azimuth and from +750 to -750 in elevation. The other type of scanner is called

Camera View Scanner. It is usually implemented by two perpendicular deflection mirrors, one

for the azimuth and the other for the elevation. This type of scanner has a fixed field of view

typically of 450x450, but it can be extended to a panoramic field of view with the aid of optional

pan and tilt bases. (Frohlich & Mettenleiter, 2004)

1.4 Technical Characteristics and Specifications of TLS

Currently there are more than 12 commercial-off-the-shelf TLS systems. With this wide

range of options, defining which is the right TLS for a project depends on the careful analysis of

the instruments specifications and the project requirements. Appendix A has a description of

commercial TLS units. Key specifications of TLS are: range, range resolution, precision,

accuracy, azimuth & elevation resolution or point spacing, LASER type & wavelength, and the

scan rate, field of view. Subsystem specific characteristics will be described first as they have an

impact on integrated system characteristics.

LASER wavelength: Even when the manufacturer won't provide the exact wavelength of

the LASER it will generally give a range such as green, red or near infrared. Some consideration

must be given to the LASER wavelength of the TLS. Visible LASERS will be best when there is

the need for water and glass penetration or when mapping wet surfaces. Most of the energy of

infrared LASERs will be absorbed by moist surfaces and the return signal will be very weak.

1.4.1 LASER Type

LASER type will be either pulsed or continuous wave. The advantage or disadvantage of

each type was discussed in the previous section.









1.4.2 LASER Class


The Class of a LASER is defined by the American National Standards Institute (ANSI)

according to the degree of hazard presented to eye safety based on a maximum permissible

exposure (MPE). The class depends on the LASER power and wavelength. TLS must be built to

meet eyes safety regulations, and the operator must be aware of what are previsions and

precautions that must be taken. Table 1-4 provides a summary of the LASER classification

scheme.

Table 1-4. LASER classification
Class Description
Class I Safe for use under all reasonably-anticipated conditions of use. It is not expected that the
MPE can be exceeded. This class may include LASERs of a higher class whose beams are
confined within a suitable enclosure so that access to LASER radiation is physically
prevented.
Class IM This LASERs produce large-diameter beams, or beams that are divergent. The MPE for a
Class IM LASER cannot normally be exceeded unless focusing or imaging optics are used
to narrow down the beam. If the beam is refocused, the Class has to be upgraded.
Class II Emits in the visible region. It is presumed that the human blink reflex will be sufficient to
prevent damaging exposure
Class IIM They emit in the visible region in the form of a large diameter or divergent beam. It is
presumed that the human blink reflex will be sufficient to prevent damaging exposure.
Class IIIR All continuous wave LASERs which may produce up to five times the emission limit for
Class 1 or class 2 LASERs. Although the MPE can be exceeded, the risk of injury is low.
The LASER can produce no more than 5 mW in the visible region.
Class IIIB They produce light of an intensity such that the MPE for eye exposure may be exceeded
and direct viewing of the beam is potentially serious.
Class IV High power (typically more than 500 mW ifcw, or 10 J/cm2 if pulsed). These are hazardous to
view at all times, may cause devastating and permanent eye damage, may have sufficient
energy to ignite materials, and may cause significant skin damage.

1.4.3 LASER Beam Divergence

The LASER beam divergence will determine the footprint area at a given range. For an

accurate mapping one will require the smallest footprint size because the LIDAR computes an

average range of the entire illuminated area. The larger the area, the more chance of slope,

reflectivity and smoothness variations affecting the range measurement.









1.4.4 Point Spacing

Point Spacing or angular resolution (azimuth and elevation) is the measure of the smallest

angular step the scanner mechanism can steer the LASER beam. In other words, it is the measure

of the angular or linear separation between adjacent LASER shots.

1.4.5 Range

Range is perhaps the most important characteristic of a TLS, and performance

specifications should include both a maximum and minimum ranging distance. Range will vary

greatly between units based on the LIDAR principle of operation and the specific design

characteristics (pulsed energy and detector sensitivity for TOF, long wavelength for PD and

baseline distance for OT).

1.4.6 Range Resolution

Range resolution refers to the ability of the TLS to distinguish between adjacent features in

the range direction. Range resolution depends on pulse width on TOF units, of phase measuring

resolution on PD, and on baseline distance and the spatial resolution of the position detector in

OT units.

1.4.7 Precision and Accuracy

Precision and accuracy are often used usually interchangeably, however, they are not the

same concept. Precision is the statistical closeness of a set of repeated measurements while

accuracy is closeness of the best estimate value obtained by the measurements to the accepted

true "value" of the measured quantity. (DMA, 1991) Baltsavias(1999) states that "ranging

precision" is inversely proportional to the square root of the signal to noise ratio and gives

expressions for the estimate of the "ranging precision" for TOF and PD LIDARs. Wehr & Lohr

(1999) present the same expressions but they refer to them as "ranging accuracy". What they are

in fact are referring is not to either ranging accuracy or precision but rather estimates of ranging









errors as a function of intrinsic electronic parameters which will have some relation to the

ranging precision and accuracy.

Before describing TLS precision or accuracy it is necessary to consider that they

reconstruct reality by measuring ranges and angles. The final coordinates of points are derived

from computations using the observed ranges and angles, and they are subject to a combination

of errors from each of the hardware subsystems as well as rounding and other computational

errors of the software subsystem. It impossible to derive closed form equations for a TLS

precision and accuracy; this parameters must be estimated through extensive laboratory testing.

The overall precision of a TLS system is the degree of repeatability of its range and

position measurements. There are two types of precision, single measurement precision and

averaged measurement precision. The single measurement precision can be understood as the

theoretical error in measuring a single point only once. Averaged measurement precision is

obtained if the system takes multiple measurements of a single point and computes a mean and

standard deviation from these observations. To determine accuracy a manufacturer has to test the

measurements derived from TLS using a higher quality data set to see how well they agree.

Manufactures usually quote the modeled accuracy, which is derived from the fitting multiple

point measurements to a primitive model. (lavarone, 2002)









CHAPTER 2
UF MOBILE TERRESTRIAL LASER SCANNING (M-TLS) SYSTEM

2.1 Evolution of the M-TLS Concept

The University of Florida (UF) was a pioneer in the application of commercial airborne

laser mapping systems to the fields of environmental and infrastructure surveying. In October,

1996, a demonstration/test project was conducted for the Florida Department of Environmental

Protection in collaboration with Optech, the Florida Department of Transportation (FDOT) and

the US Geological Survey Center for Coastal Geology. During project LASER (Laser Swath-

mapping Evaluation and Resurvey) more than three hundred kilometers of beaches (Mexico

Beach, FL, to the western tip of Perdido Key, AL) and a portion of Interstate 10 were mapped

using an Optech Inc. ALTM 1020 laser ranging system. (Carter & Shrestha, 1997). Shortly after,

in 1998 UF in conjunction with the Florida International University (FIU) purchased its first

LIDAR mapping unit, an Optech 1020 ALTM.

In early 2001, the University of Florida (UF), the National Geodetic Survey (NGS), the

FAA, and Optech, Inc. conducted a field test to explore using an Airborne Laser Scanner for

Detecting Airport Obstructions at the Gainesville Regional Airport. (Parrish et al., 2005) During

this research UF personnel tested a preproduction prototype of an Optech's tripod mounted

LIDAR unit named "Intelligent Laser Ranging Imaging System" (ILRIS-3D).

After the tragic attacks on September 11, 2001, the Office of the Deputy Under-secretary

of Defense approached the US Army's Joint Precision Strike Demonstration (JPSD) to inquire

about specific technology capabilities to aid the surveying the NY ground zero. JPSD

approached Optech and the University of Florida Geosensing Systems Engineering center for

personnel and equipment support. A multi institutional group was established and laser









surveying began on September 18. Besides the airborne unit, two ILRIS-3D systems were

deployed to map rubble piles and damaged building structures (Kern, 2001).

These experiences with the Terrestrial Laser Scanner made the UF researches quickly

realize the potential of the technology as a valuable complement to the airborne laser unit.

During 2002 the UF Geosensing Systems Engineering (GSE) division acquired and fully tested

an ILRIS-3D unit. The results of the tests are thoroughly discussed in the 2002 Master thesis

"Applications of Laser Scanning and Imaging Systems" by GSE student Devin Robert Drake.

In 2002 UF prepared a proposal for the Florida Department of Transportation of a Mobile

Laser Surveying System (MOBLASS). Potential applications of the system were identified to

include the precise positioning, the continued evaluation and documentation of the primary

components of the transportation infrastructure (center lines, guardrails, signs, bridges

maintenance facilities). The proposal consisted on a twelve passenger van equipped with an

ILRIS 3D unit mounted on a telescopic pan tilt base, a GPS array for positional and azimuth

determination, power leveling and stabilizing units, an auxiliary power generator, an operator

control console with display, a PC for data reduction and analysis, a wireless data link for

realtime transmission to the operational center and upgrade capabilities for an IMU and other

types of sensors (digital imaging, hyperspectral, ground penetrating radar). At that point in time

technology just permitted static mapping and the proposal was developed as a Stop-Map-and -

Go system, unfortunately FDOT decided not to fund the development of the MOBLASS. Figure

2-1 shows a survey vehicle developed around a concept similar to MOBLASS.

In 2005 the UF Geosensing Systems Engineering division decided to develop the

MOBLASS concept with its own resources. The system name was changed to M-TLS for

Mobile Terrestrial Laser Scanning system and is based on a new version of the ILRIS 3D system









which enables the interfacing with an Optech produced Pan Tilt base. Other mayor change in the

concept is that the vehicle is now a 4x4 truck which extends the terrain operation capabilities and

allows an increased instrumental payload. The M-TLS is a unique tool that enables the

acquisition of high density point clouds from an advantageous terrestrial geometry, being a very

valuable complement for Airborne Laser Scanner data sets.











-'I



Figure 2-1. Survey vehicle with a similar concept to MOBLASS.

2.2 M-TLS Subsystems

2.2.1 LIDAR Unit

The core of the M-TLS is a commercial 2-axis ground based laser scanner which is

integrated to a mobile telescoping, rotating and tilting platform which provide up to 6 degrees of

freedom for performing scanning operations. A scanner built-in 6 megapixel digital camera and a

digital video camera provide the M-TLS still and video imagining capability.

The laser scanner is an Optech ILRIS-36D, which is capable of generating XYZ with laser

intensity or RGB textured point clouds in a range from 3 m to 1500 m for a target with an 80%

reflectivity or 3 m to 350 m to targets with a 4% reflectivity. The laser operates at a wavelength

of 1535 nm, with a pulse width less than 10 ns and energy of less than 10 [tjoules. The sample

separation can be adjusted down to 0.001150, and the scanning speed is 2,000 points per second.









Currently the ILRIS is only capable of performing its scanning and mapping operations on

a static mode. In a near future Optech will release a unit with "on-the-move" mapping

capabilities, at that point is expected that the M-TLS will perform dynamic mapping operations.


D Digital Camera
Controller ........ O. Cable
Video B.E.
Display A77X Axis Scanner
D isplayM n ....... Fi ........................ :anner

Main
CPU
C PX V : ................ ................
SBandpass Fi .- .. ..... .....
USB
USB "s'MIIlnGaAs APD
Storage /
N TIM Discrm. *Receiver Y Axis Scanner

Network Y Axis Driver
Interface Analog I/O X Axis Driver
Digital I/O -

Figure 2-2. OPTECH ILRIS block diagram.

Figure 2-2 shows a block diagram of the ILRIS unit. At a very high level the internal

operation of the ILRIS is quite simple. A central microprocessor controls the different

subsystems, collect, analyzes and displays data and information. The computer commands a laser

controller to fire a pulse, the laser beam generated by a Nd YAG laser is passed thru a non linear

crystal that shifts its natural frequency from 1064 nm to 1535 nm. Then the laser beam pass thru

a beam expander and a small amount of photons are diverted thru a fiber optic cable to start the

time of flight timer.

After the beam is expanded it passes through an optical element that reflects the returning

beam to the detector. The beam is then deflected by the vertical and horizontal scanning mirror

to the target. The position of each of the mirrors is controlled by the scanner axis drivers and the









central computer. Most of the targets have quasi-Lambertian reflective properties and will reflect

the laser beam as a distorted waveform, part of that waveform will return in the same optical path

of the outgoing beam. The incoming waveform will be reflected by the scanning mirrors to a

fixed parabolic mirror that focus the waveform on to the detector. Prior to entering the detector a

narrowband optical filter centered at the 1535 nm limits the noise entering the system. When the

returning laser photons arrive at the detector, which is an indium gallium arsenide (InGaAs)

Avalanche Photo Diode (APD), a voltage is generated between its terminals. The output voltage

is read by an A/D converter where the signal is digitized and sent to a Constant Fraction

Discriminator (CFD). The CFD is an electronic device designed to produce accurate timing

information from signals of varying amplitudes but the same rise time. CFD usually achieve this

by splitting the input signal, attenuating half of it and delaying the other half, then feeding the

two halves into a fast comparator with the delayed input inverted. By doing this CFD is capable

of triggering a timing signal at a constant fraction of the input amplitude. The CFD trigger is fed

into the precise Time Interval Meter (TIM) which was original started by the outgoing pulse feed

to the detector thru the optical fiber. The TIM computed the time difference between the

outgoing and incoming pulse thus determining the 2 way time-of-flight. The computer then

calculates the range to the target records it on it internal memory and commands the emission of

a new pulse and the entire process is repeated until the defined area of interest is scanned.

The block diagram shows additional subcomponents of the ILRIS unit. On-board 6-

megapixel CMOS digital camera is used to provide the operator with a low scan rate video and

high resolution stills of the scan area. The Camera is boresight calibrated with the LIDAR to

provide red, green and blue channels to each laser point. The output from the camera is also

projected with other control information to a 5.5 x 4 LCD viewfinder. On board data storage









is done on a conventional USB jump drive. The controlling and commanding of the ILRIS is

performed thru proprietary software called "Controller" which runs from a computer or pocket

pc. To provide connectivity to the PC the ILRIS has two network interfaces devices: a Ethernet

interface card (IEEE 802.3) and a wireless network card (IEEE 802.11). Power to the ILRIS is

provided from a 28 V DC battery pack or from a 120 VAC/28 VDC power adapter.

2.2.2 Vehicle

The vehicle selected for the M-TLS system is a Ford F250 4x4 crew cab long bed truck.

The truck has undergone several modifications including the installation of a steel frame for

mounting the telescopic lift and the installation of four electrical jacks which will enable

automatic vehicle stabilization and leveling. The vehicle 4x4 capability allows the execution of

off-road mapping projects, its 1600+ kg of cargo capacity allows the loading of the 374 kg lift

with ample capacity for mapping, positioning and auxiliary equipment. Figure 2-3 shows the M-

TLS truck with the lift deployed at half height.


Figure 2-3. Mobile Terrestrial Laser Scanning (M-TLS) system: truck, lift and ILRIS.









2.2.3 Lift

The lift is an AC powered JLG Push-Around Vertical Lift model AM 25. This lift has a

stowed height of 1.97 m and once deployed it extends to 9.45 m. The truck was modified by

bolting a steel frame directly to the back chassis, the lift then rests over the steel frame. An

instrument aluminum frame was constructed to support the ILRIS, its power supply and an

electronic tilt meter. The instrument frame is mounted on the top of the lift. With all these

provision the ILRIS scanner can be lifted to a vantage point of up to 12 meters above ground

level.

2.2.4 Power Subsystem

The M-TLS have components that require both AC and DC power. To supply the power

requirements the M-TLS system has a Briggs & Stratton 6200 Watts electric start generator and

a 12V DC battery bank. The electric generator is used when the lift needs to raised or lowered

and to charge the battery bank. DC power is directly available from the battery bank and low

power AC devices can be fed thru a DC to AC power inverter from the battery bank. Figure 2-4

shows the power plant installed on the truck bed.


Figure 2-4. M-TLS deployed in Georgia, showing the power generator and held computer on the
truck bed.









2.2.5 Pan Tilt Base

Currently there are two pan & tilt bases available on the M-TLS system. The first one is an

Optech manufactured base that is connected to the ILRIS 3D unit and controlled from the ILRIS

control software. This base allows a complete 3600 rotation in azimuth and 700 in elevation, with

the 400x40 ILRIS's field of view the tilt base permits a -20 to 900or a -900 to -200 elevation

coverage. The advantage of using this base is that the scan data can be automatically de-rotated

and de-tilted in the Parsing process, yielding a complete coherent data set. The disadvantage is

that when the ILRIS is powered up and it detects that is connected to a pan tilt base it will

capture a 3600 panoramic picture composed of 10 individual digital camera frames and this can

be extremely time consuming if the user is just interested in scanning a narrow field. To

overcome this disadvantage a second pan tilt base is available; the base was manufactured by

QuickSet International and is designed for the operation of surveillance cameras. It has a loading

capacity of 40 kg and a rotation range of 217.50 in azimuth and 900 in elevation. This pan tilt

base can be controlled from a PC or from an analog joystick console. Figure 2-5 shows the two

available pan tilt bases for the M-TLS.

















Figure 2-5. M-TLS pan tilt bases. A) Optech 360 base. B) Quickset QPT pan tilt base.









2.2.6 Video Camera

The MTLS is equipped with a Samsung SCC-C4201 high resolution color video camera.

The camera has a built in X22 optical lens and an X10 digital zoom to provide an effective,

enhanced focal length of 3.6 to 79.2mm. The 410,000 pixels CCD (811 x 508) outputs NTSC

with 480 Horizontal Lines and 350 Vertical Lines. The video provides the operator with wide

and narrow views of the scan project area. The output can be directly viewed on a TV screen or

on the computer thru a video capture card. The video camera can also be used as a surveying and

mapping tool during operation of the M-TLS dynamic mode. The video camera is mounted on

top of the ILRIS unit by means of a special housing as shown in Figure 2-6.

2.2.7 On Board PC

Currently only data collection is done onboard. For this purpose a laptop containing the

ILRIS controller software is used. The ILRIS to PC connection is done by the wireless peer to

peer network. Data is stored in both the ILRIS USB disk and on the laptop hard drive. On a

future a central computer that will control and record the data from the ILRIS, INS and cameras

will be installed.

2.2.8 GPS

At this point in time the ILRIS unit can not do direct georeferencing. In the static mode,

when a geo-referenced dataset is desired the use of GPS control points is required. This is

usually done by surveying at least three control points well distributed both vertically and

horizontally on the area to be scanned. An additional control point can be located by installing a

GPS antenna on top of the ILRIS unit, by adding the proper XYZ offsets the GPS coordinates of

that antenna reference point can be translated to the ILRIS coordinate origin. Figure 2-6 shows

the provisions for the installation of a medium size L1/L2 marine antenna on the ILRIS unit. For

the purpose of surveying the control points a set of geodetic grade GPS receiver are available










which include units from the ASHTECH models Z-Xtreme, Z-Surveyor and Z-12. Ashtech

L1/L2 Choke Ring and marine L1/L2 model 700700 are used to collect the GPS signals. Figure

2-7 shows an array of GPS stations used for geo-referencing of M-TLS data set during the

forestry experiment.



I. .........:






















Figure 2-6. The ILRIS unit with on axis video camera and GPS antenna.



.! T7..


Figure 2-7. Array of geodetic quality GPS base stations used for geo-referencing.









2.2.9 Tiltmeter

An electronic tilt meter is installed under the instrument frame on the lift to measure the

vibrations and motions to which the ILRIS is subjected and to ensure that during scanning

operations the platform remains as steady as possible. The red ellipse on Figure 2-8 marks the

installation position of the Tiltmeter unit.





















Figure 2-8. Installation of the tiltmeter unit on the instrument frame.

2.2.10 INS

When Optech releases the new generation of ILRIS with on the move scanning capability,

the M-TLS will be capable of performing mapping operations in the dynamic mode. At this point

the installation of an Integrated Navigation System must be performed. This can be in the form

of a canned solution like the Applanix POS LV or with an in-house developed INS. The

Applanix POS LV system includes a single frequency two antenna survey grade GPS array and

an Inertial Measuring Unit. The IMU grade depends of the clearance obtained by the US

Department of State regarding the International Traffic in Arms Regulations.









CHAPTER 3
WORKFLOW OF M-TLS OPERATIONS

3.1 Data Collection

Data collection design is the first step in M-TLS operations. There are no written rules on

how to perform a data collection, it depends on very specific details of the project and the

experience of the operator. Data collection design starts with an analysis of the requirements of

area of interest; desired resolution or laser point spacing; laser return intensity or RGB texture;

data set reference (sensor XYZ or geo-referenced to a particular datum) and accessibility to the

scan area. With these inputs the operator defines the data set acquisition strategy that includes the

selection of the scanning geometries (number and orientation of the scans) and the design of the

GPS Control Points Network if the data set needs to be referenced to a particular datum rather

than the natural XYZ sensor frame of reference.

Collection is performed with the Optech proprietary software "Controller". The controller

main screen displays a color image which covers the 400x40 scannable field of view. The

operator then selects a Region of Interest (ROI) to be scanned and based on a preliminary range

acquisition, adjusts the spot spacing in angular or linear units. After setting the data storage

directory the scan can be initiated. When using the Optech pan and tilt base the ILRIS will

acquire a 3600x40 panorama consisting of 10 overlapping still frames. The complete panorama

will be displayed in Controller window and the user can then set the ROI to be scanned. Figure

3-1 presents a screen capture of the Controller software.

For particular projects several setups are required. In some cases a combination of

panoramic and frame scans have to be performed. The integrated M-TLS system provides up to 6

degrees of freedom (4 rotations and 2 translations) for performing scanning operations that

guarantee that all possible facets of the object will be mapped under the most favorable









conditions. Figure 3-2 shows photographs of different configurations of the M-TLS used during


data collections.


Figure 3-1. Screen capture of the ILRIS unit controller software during scanning operations.


Figure 3-2. Different configurations of the M-TLS system during several data collection projects.
A) Forest scanning B) Soil roughness experiment, a microwave radiometer can be
seen behind the M-TLS. C) Beach erosion hot spot mapping in St. Augustine Fl.

3.2 Data Parsing

At collection the data is stored in a binary format and includes sensor orientation

parameters and range for each measurements made. The raw data needs to be converted into a


~P~8..........









position in a coherent 3D frame of reference. Additional information can be texture or color

information from the intensity of the reflected laser signal or from a coregistered imaging sensor.

This first step of processing is called Parsing and it is performed by the proprietary software

"PARSER". With Parser the raw data is converted to any of the known point cloud formats that

can be read by most LIDAR processing software for further manipulation and analysis.

Figure 3-3 shows a screen capture of the Parser software. The large screen shows the

digital image of the scan surface, the red box encloses the selected region of interest. The Parser

Settings tab allows the user to personalize the conversion process. There are options for the

output file format, the choice of using the digital image to provide RGB color channels for each

of the scanned points, smoothing of the range shots, corrections based on the Pan-Tilt orientation

and an option for the positioning of the measurement origin.


Figure 3-3. Screen capture of the Parser software during the setup of parsing settings.









The most common output selected is the ASCII XYZ File which converts the range,

azimuth and elevation into a coordinate in a right hand Cartesian frame with the origin at the

LIDAR reference point. Additional information can include a normalized 8 bit laser return

intensity or RGB channels. This format is easily uploaded into point cloud processing software

like QT Modeler, Terra Scan, Polyworks or Matlab for a customized analysis.

3.3 Data Manipulation and Information Extraction

Once the data is parsed is ready for uploading into specialized software that allow its

manipulation, analysis and information extraction. The information extraction from a M-TLS is

the final goal, it is a customized process for each application. Some of these specialized

techniques are discussed on an application to application basis on chapters 4 to 7. However, there

are several typical operations that can be performed on LIDAR point cloud data; these are

visualization, transformations, segmentation, classification, filtering, gridding and specialized

mathematical operations.

3.3.1 Visualization

The first thing that is done to a LIDAR data set is to look at it. Visualization is the most

basic operation; however, a good visualization allows the analyst to asses the quality of the data

set, it enables the planning and control of different processing schemes and finally will provide

the presentation of the final product. Most LIDAR processing software will have a graphical

interface that will render the numerical point cloud into an image, but there is a great range of

options and functionalities that will vary among the different options. The simplest visualization

will plot all the points with a single color and size, and the operations of Zoom, Rotate, and

Navigate will be available. More advanced software will render each LIDAR point according to

other characteristics, it can be brightness coded according to the laser return intensity, or RGB

textured if the point cloud was coregistered with a digital image, it can also be color coded









according to elevation, range, class or any other attribute contained in the point cloud structure.

Some software will allow the user to toggle between the rendering of the point cloud and the

rendering of a Triangulated Irregular Network (TIN) or Digital Elevation Model generated from

it, as shown in Figure 3-4.

















Figure 3-4. Visualization of LIDAR data. A) As a point cloud. B) As a DEM.

3.3.2 Single Point Selection

An important functionality of visualization software is the one that allows the user to

manually do single point selection. This is to navigate through the point cloud using the zoom

and rotate controls to pick out single points from the cloud.

3.3.3 Measurements

The ability to precisely select points from the clouds allows the analyst to make

measurements such as distances between points, and angles between lines connecting the points.

3.3.4 Primitive Fitting

After selecting a series of points is possible to perform a primitive fitting operation.

Primitive fitting is the application of the least square methodology to compute the spatial

parameters that define simple geometric figures or volumes such as lines, circles, planes,

spheres, cones. Primitive fitting allows computing the modeling accuracy of TLS. Imagine that









there is a sphere with a known radius. After the sphere is scanned a best fitting sphere is

determined for the point cloud using least squares to minimize the residuals. The "fitted" or

"modeled" sphere radius can then be compared to the known radius and the degree of agreement

between both provides an estimate of the modeling accuracy. Figure 3-5 illustrate the entire

process from point selection to the fitting of the sphere.


figure 3-3. rimitive rating process inustratea. A) roint cloua. 1) icKming points irom me
sphere surface. C) Fitted sphere based on the picked points.

3.3.5 Generating Cross Sections

An important visualization tool is the selection of a particular baseline and the generation

of a cross sectional view of the point cloud at that baseline.


Figure 3-6. Cross section generation from the point cloud.









3.3.6 Transformations

There are countless transformations that can be applied to the point clouds, a few of the

most frequent are described following:

3.3.6.1 Rotations and translations

Simple Transformation includes the translation or rotation of the entire point cloud on one

or more of the coordinate axes.

3.3.6.2 Cropping

When an object is scanned there are always points that do not belong to the volume of

interest. Cropping allows the creation of a point cloud with only the elements that falls within the

3D space of interest.

3.3.6.3 Merging

Point Cloud Merging is performed when several point clouds of the same object were

collected from different angles or positions each having its own coordinate frame and there is the

need to convert all of them into a single spatial coherent point cloud. Merging is performed by

setting one point cloud as the base reference frame, and then common points or common

primitives are identified between the base and the source point cloud. From the common points a

3D rotation and translation transformation is computed using least squares adjustment. Then the

transformation is applied to the source point cloud to change its coordinate frame to the base

reference frame. Figure 3-7 illustrates a merging operation between two point clouds color coded

as white and pink that were obtained from different scan angles, using the common points

method the pink point cloud was rotated to the white coordinate system to produce a single

coherent data set.

With the Polyworks IMspect and IMAlign software the common points can be manually

selected and then the transformation process is done automatically. Another option is to use









Polyworks IMspect to find the common point coordinates in both the source and the target

frames create text files with those coordinates and then use TerraScan transformation modules to

compute the parameters of the transformation and then manually apply it to the source point

cloud. The advantage of this latter option is that TerraScan provides a residual analysis of the

transformation based on the common points and this give the user an idea of the quality of the

transformation.









Figure 3-7. Point cloud merging example.

3.3.6.4 Geo-referencing

A transformation in which a point cloud with coordinates in arbitrary sensor space is

converted into a geodetic coordinate frame is called geo-referencing. This operation has to be

performed when absolute measurements have to be made or when the terrestrial data set will be

merged or compared to an airborne data set. Similar to a merging operation, in geo-referencing

there has to be a minimum of 3 non-collinear points for which coordinates on both sensor and

geodetic frames are known. Based on that set of coordinates the parameters of a 3D rotation and

translation transformation are computed. That transformation is then applied to the entire point

cloud and as a result the data set is fixed to the specific geodetic frame. The accuracy of the geo-

referencing depends primarily of the quality of the GPS observations, the vertical and horizontal

strength of the control points network and the determination of the XYZ coordinates of the

control points from the original point cloud. Geo-referencing can be performed using both

Polyworks and TerraScan.









3.3.7 Segmentation, Classification and Filtering

Another important set of operations performed over the point clouds are the ones that

allow performing segmentation, classification and filtering of the points.

3.3.7.1 Segmentation

Segmentation refers to the operation that will segment or segregate points into different

groups based on characteristics without knowledge of what they really are. An example of

segmentation could be the separation of points, based on intensity values, into low intensity,

medium intensity and high intensity. Under this segmentation scheme points in each group will

not necessarily share common spatial characteristics.

3.3.7.2 Classification

Classification implies the separation of points into different groups or classes defined by an

intrinsic or natural characteristic. An example of classification is the separation of the points into

vegetation, building or ground classes; each of these groups implies the knowledge of its nature.

3.3.7.3 Filtering

Filtering is the removal of a set of points from the clouds based on either a segmentation or

classification scheme. An example of a segmentation scheme based filter could be the removal of

points that are below a certain height value, without considering its nature (i.e. ground or low

vegetation). A classification filter could be one that removes vegetation from an urban scene on

which only brick and glass is wanted.

3.3.8 Gridding

A scanner point cloud by nature is an irregularly space data set. The process of converting

the point cloud into a regularly spaced data set by means of interpolation is called gridding.

Gridding allows the analyst to observe subtle features in the data set. There are many different

gridding algorithms the more common are Nearest Neighbor, Inverse Distance Weighting,









Triangulation with Linear Interpolation, and Kriging. The regular nature of the grid allows the

analyst to perform many mathematical operations such as areas and volumes computations, grid

algebra, grid calculus, differentiation, gradients, grid comparison, as well as image processing

operations. Gridding can be performed using specialized software such as Golden software

"Surfer" or with built-in routines in Matlab. Figure 3-8 shows a regular grid surface model B)

generated from the irregular spaced point cloud.


Figure 3-8. Gridding operations, from point cloud to grid. A) Point cloud to B) Surface Mod,

3.3.9 Advanced Mathematical Operations

The operations discussed so far are commonly performed by the LIDAR analyst using

canned algorithms in commercial software packages. However, some applications require









advanced or specialized techniques that must be custom programmed in programming languages

such as Visual C, Visual Basic or using higher level math tools such as Matlab or IDL. Examples

of these advanced mathematical operations may include:

* Transformations from space to the frequency domain using the Discrete Fourier Transform
or with the Discrete Wavelet Transform.

* The use of spin images to represent objects from a 3D dataset in a single 2D image.

* The application of advance image processing techniques and operations such as edge
detection or morphological operations to a gridded dataset.

Figure 3-9 shows an example of and advanced mathematical operation. Were a discrete 2D

Fast Fourier Transform was used to extract the stronger periodic components of the real terrain

and then a 3D surface was generated from those components.

A C











B D









Figure 3-9. Examples of advanced mathematical operations in the processing of Airborne
LIDAR data. Digital elevation models of A) Gabilan Mesa, Ca. and B) South Fork
Eel river, Ca. C) and D) show visual representation of the most strongly periodic
component of each landscape. (Perron, 2006)









CHAPTER 4
TESTED APPLICATIONS OF M-TLS

4.1 Common Applications of TLS

TLS units are marketed by the manufacturers mainly as surveying tools for engineering

applications. These engineering applications include as-built-surveys, crime scene and traffic

accident investigation, mine operations planning and supervision, transportation infrastructure

mapping, bridge loading analysis, building damage assessment and urban modeling. The reason

for this biased marketing is that TLS are expensive instruments that are often affordable only by

companies that do large projects with huge capital investments, where the savings of time over

classical mapping techniques are even greater than the cost of the TLS instrumentation. But even

with that biased marketing over the last 5 years, some articles about scientific applications of

TLS have been published in scientific journals or presented in professional meetings.

One of the first presentations reporting the application of TLS technology to scientific

research was that of the University of Texas Bureau of Economic Geology "3-Dimensional

Digital Outcrop Data Collection and Analysis Using Eye-safe Laser (LIDAR) Technology"

presented on the 2002 convention of the American Association of Petroleum Geologists (Bellian

et al., 2002). In 2003 a paper published in the International Society for Optical Engineering

(SPIE) Optical Metrology for Arts and Multimedia journal entitled "High-resolution laser radar

for 3D imaging in artwork cataloging, reproduction, and restoration" introduced the application

of TLS for cultural heritage preservation. With respect to forestry one of the first articles

published was "Using airborne and ground based ranging LIDAR to measure canopy structure in

Australian forests" published on the Canadian Journal on Remote Sensing vol. 29, 2003

(Hopkinson et al., 2004).









These papers were the groundbreakers of the scientific applications of TLS. It is clear that

TLS has a great potential to contribute to several fields of science that require precise spatial

measurements. It is just a matter of making the technology available to researchers and providing

them with technical support on the data processing and information extraction. Just as Airborne

Laser Mapping proved to be a valuable tool, TLS will be accepted if it proves to be better that

conventional data gathering techniques by providing higher quality and quantity of data faster

and cheaper. Several applications were tested under this philosophy, a subset of which are briefly

described in the following sections. These sections present a "Traditional" vs "Alternative"

methods of collecting spatial field measurements. Subsequently, three of these applications will

be thoroughly discussed in the next chapters, and include quantitative comparisons with the

traditional field methods techniques.

4.2 Paleontology

Paleontology and archeology field techniques require extensive digging and constant

measurements. Usually at a dig site, before digging operations begin, a regular grid is

established. The positions of the fossils or artifacts that are discovered are carefully measured

and recorded by photography. The current field methods are very time consuming can really

slow the dig progress specially when a very important finds such as articulate skeletons or

unique artifacts are discovered. TLS systems can be used in these type of cases to provide both

photographic and 3D measurements. Figure 4.1 illustrate both the traditional and the alternative

approach to obtain measurements at dig sites. Part A shows the traditional way were stakes and

cordas are used to define the grid, while part B shows a TLS measuring and recording a dig

site.Digital terrain models from TLS data can provide additional information for monitoring

progress such as dig volume and cleared surface.










i .0
A_


. B


Figure 4-1. Traditional and alternative methods for measuring and recording spatial information
in archeological and paleontological sites. A) Traditional grid. B) TLS mapping.

The paleontological application was tested at a Florida Museum of Natural History dig site

located near Haile, Newberry; it is identified by Florida Museum as H7G. Figure 4-lb shows the

ILRIS mounted on a tripod at that dig site. The site is a large sinkhole in a limestone quarry,

containing fossil bones and teeth of about 40 different kinds of freshwater and land animals

dated at about 2 million years old (Pliocene Epoch). Some fossils are preserved as intact

skeletons but mainly isolated bones or teeth constitute the major finds. The study objectives

were: 1) to establish a macro Geo-referenced 3D model of the dig site from several large scale

ILRIS scans; 2) to obtain high density point clouds of small patches of the dig showing the

fossils matched to the macro model and 3) if an interesting specimen was discovered the

development of a multi angle high density point cloud model would be developed and embedded

into the macro model.

Figure 4-2 shows a point cloud rendering of the macro model of the dig site. Figure 4-3

shows RGB textured renderings of high resolution point clouds obtained from a small patch of

the dig site and Figure 4-4 presents before and after surface models generated from the point

clouds. These were later used to determine the dig volume (- 2.0 m3) after one day of

excavation.






























Figure 4-2. Rendering of the dig site point cloud macro model showing laser return intensity.


Figure 4-3. RGB textured renderings of high resolution point clouds.













Difference between Models


SNet Volume Difference -2.0 m3





Second'Surface Model .,



Figure 4-4. 3D surface grids used to compute the volume of dirt extracted in one day.

4.3 Structural Geology

Geology is a discipline that relies extensively on mapping for the spatial recording of

topography, crustal elements like faults and folds, stratigraphy and many other features. Plane

table mapping is a traditional geological mapping technique, on which simple tools (a flat

leveled table, an angular measuring device and a scale) are used to create large scale (1:120)

maps. Plane table was replaced by modern surveying equipment such as total stations and

computer aided design (CAD) software. However, creating maps using the plane table or a total

station are highly time consuming processes. Figure 4-5 shows the traditional table mapping

technique and the alternative terrestrial laser scanning technique.





















Figure 4-5. Traditional and alternative ways to perform geological field mapping. A) Traditional
plane table mapping technique. B) Moder alternative mapping using TLS.

TLS are ideal for creating geological maps for the study of geomorphology, stratigraphy

and structural geomechanics as described by University of Texas Bureau of Economic Geology

(Bellian et al., 2002). The tested application of TLS to structural Geology was in support of

NCALM PI Stephen Martel of the Department of Geology and Geophysics, University of

Hawai'i at Manoa. Dr Martel has been developing theoretical models on the effect of

topographic curvature on near-surface stresses and the creation of sheeting joints. Sheeting joints

are opening mode rock fractures that form subparallel to the topographic surface, develop to

depths of at least 100 m and they occur mainly on regions where the topography is convex

(Martel, 2006). The geology of Yosemite National Park in California exhibits vast areas with

exposed sheeting joints; TLS mapping was performed on the Tuolumne quarry located along the

Tioga road. The objectives of the project were 1) to map the sheeting joints on the exposed wall

of the quarry on a 3D geodetic space and, 2) to the sheeting joints orientation to the topographic

curvature of the area obtained from ALSM and the mechanical stress to which the formation is

subjected.

Data was collected on September 22 and 23, 2006. The dataset consisted of 7 overlapping

scans and vector observations for 6 reference GPS base stations that were used for the geo-

referencing of the data sets. The binary scan files were parsed to generate the XYZ & Laser









Intensity and XYZ & RGB ASCII text files. Using Innovmetric Polyworks IMInspect module

common points were identified in the overlapping scans to perform the merging operations. The

coordinates of common points were input to Terrasolid Terrascan to compute the solid 3D

translation and rotation transformation parameters. After the transformation parameters were

computed, the six sets of point clouds were transformed to the central scan coordinate system

and a single point cloud was generated.

Using Polyworks IMInspect module, the GPS control points were identified in the merged

point cloud and their sensor XYZ coordinates were determined. The GPS observation files were

processed using the NGS Online Positioning User Service (OPUS)

(http://www.ngs.noaa.gov/OPUS/) and the GPS coordinates for the control points were

determined. The standard deviation of the coordinate components varied between 0.014 to 0.789

meters, with a mean of 0.219 meters. With the sensor and UTM coordinates of the GPS control

points; Terrascan was used to compute the solid 3D translation and rotation transformation

parameters for the data set. The RMS of the residuals of the transformation on the control points

were 0.253919157, 0.156471707, 0.423406924 meters for the Easting, Northing and Elevation

components. The merged data set was geo-referenced and transformed to a UTM zone 11

(NAD_83) point cloud.

Using Applied Imagery QT Modeler, intensity and RGB textured images were generated

from the point cloud. In Figure 4-6 renderings of the point clouds from top and front view are

presented. Using Terrascan, the point cloud was broken into tiles that later were imported to

Golden Software Surfer to produce regular 2cm grids of the vertical walls. The triangulation with

linear interpolation algorithm was selected for the grid creation. From the grids, shaded relief

images were produced; Figure 4-7 shows one of such images where the layered sheeting joints










are clearly seen. The images rendered from the grids allows to determine the 3D spatial

orientation of the joints with respect to the surrounding topographic landscape.


A B


Sigure 4-0. ueo-reterencec point cloud rendering ot the luolumne quarry. A) lop view. t3)
Front view.


2594


2592
c4)
(n
' 2590


S2588


a 2586
a-

2584


UTM Zone 11 -IxEastina metersl
Figure 4-7. Shaded relief image from a gridded model of the south wall of the quarry.









4.4 Wildlife Management Conservation

Typical wildlife management and conservation activities include the catch and release of

specimens for measuring, weighing and biological sample collection. Under certain

circumstances where there is a high risk of injury to wildlife personnel, when it is desired that the

specimen not undergo the stress of capture or when just spatial measurements are needed a TLS

may be an efficient alternative to the capture, measure and release method. In this experiment an

alligator was scanned from a safe distance of twenty meters without perturbing the specimen.

The results prove that is feasible to obtain accurate measurements of several dimensions such as

length and thorax diameter. Volumetric models can also be created from the data, which provide

a wealth of information for time series analysis. Figure 4-8 shows the traditional way of

measuring alligators or crocodiles, it also shows the rendering of a point cloud obtained by TLS

of an alligator. From the point cloud the alligator length was measured.

A B










Figure 4-8. Methods of measuring alligators and crocodiles. A) Traditional method. B)
Alternative method from TLS point clouds.

4.5 Coastal Morphology

The state of Florida has over 32000 kilometers of tidal shoreline of which more that

960 kilometers are beaches. The activities that are generated around these beaches, such as

tourism, constitute a great source of income for the state. A large portion of the low-lying sandy

beaches and dunes along the Atlantic coast are subject to modification by high surf generated by









northeasterly winds or by catastrophic phenomena as tropical storms and hurricanes. Traditional

methods of data collection and map generation for beach profile change studies include

differential leveling, traversing, static and kinematic GPS and aerial photogrammetry. These

techniques are not only costly, time-consuming, and labor-intensive but also have poor spatial

resolution (Shrestha, et al. 2005). Current Florida Department of Environmental Protection

(FDEP) standards for beach profile topographic surveying require cross shore transects at Bureau

of Beaches and Coastal Systems (BBCS) reference points, which are approximately 1,000 feet

apart along shore, with a collection interval not to exceed 25 feet. and at all grade breaks and

attributed items along the profile sufficient to accurately describe the topography at the profile

locations (BBCC, 2004). These techniques do not have the spatial or temporal resolution

required to precisely quantify and study the processes of beach erosion, especially on erosion

hotspots. Part A OF Figure 4-9 illustrate the use of RTK GPS to generate beach transects and

part B shows the UF M-TLS system used to generate beach surface maps.

.. .. ... ..













Figure 4-9. Methods for generating beach profiles. A) Using Real Time Kinematic (RTK) GPS.
B) Alternative method using the M-TLS.

The M-TLS was used to monitor an erosion hot spot located near the St. Augustine pier,

Fl. at high spatial (cm level) and temporal (biweekly) resolution. The methodology employed for

data collection and analysis as well as the results are presented in Chapter 5.









4.6 Soil Science

In soil science; active and passive microwave remote sensing techniques are applied to the

derivation of soil parameters such as temperature and moisture. In order to derive soil moisture

from radar sensors it is necessary to have a priori knowledge of the soil surface roughness.

Traditionally soil roughness has been characterized as a single scale process obtained from 2D

profiles and parameterized by the root mean square (RMS) of the height (s), the correlation

length (1) and autocorrelation function (l(h)).

The traditional methods of obtaining the soil profiles are the Needle-like Profiler and the

Mesh Board. A complete description of these mechanical profilers and the data collection

procedures is provided by Mattia et al. (2003). The main disadvantage of these mechanical

methods is that they tend to disturb the surface that is under study. Mesh Boards have to be

hammered into the soil, while Needle-like Profilers tend to penetrate into the surface yielding

noisy measurements of the heights. Modern methods are aimed to not disturb the surface; they

are non-contact instruments. These instruments include laser profilers, optical imagers and

acoustic backscatter instruments (Mattia et al., 2003; Zribi et al., 2000; Oelze et al., 2003).

Figure 4.10 illustrate the meshboard and alternate method for deriving soil roughness metrics.


A B










Figure 4-10. Methods for deriving soil roughness metrics. A) Traditional way using the mesh
board. B) Alternative method using TLS to create 3D maps of the soil surface.









The M-TLS is an excellent tool for soil roughness measurements because it has the

capability of producing terrain models at sub-centimeter scales with the additional advantage

over traditional instruments and methods that it provides a complete surface digitizing, rather

than digitized line profiles. The M-TLS was used to collect data in an experimental plot at the

University of Florida Plant Science Research and Education Center at Citra, Florida and at a

commercial plantation plot near Hastings, Fl. A detailed description of the data reduction,

analysis and result is presented in Chapter 6.

4.7 Forestry

The estimation of forest structure and volume has for a long time been of great interest to

the scientific community because of its ecological and economic importance. Traditional

methods of performing these measurements and estimations even over a small plot of forest

require a great amount of man power and many hours of field work. These methods require the

manual measurement of tree height, stem diameter at breast height (dbh), stem location and stem

density. Figure 4-11 part A shows the traditional way of measuring dbh, part B illustrate the use

of a Biltmore stick to estimate tree height, part C show the alternative method of using TLS to

generate point clouds from which dbh and tree height can be derived.

___ r__I _'_g B


gure 4-11. ivietnoas tor estimating forestry metrics. A) Iraaitional metnoa or measuring aon.
B) Traditional method for estimating tree height. C) Alternative method using the M-
TLS to create 3D maps of the forest.









Since the early 1990s Airborne Laser Mapping (ALM) has been used to determine forest

metrics. ALM provides a large spatial coverage with very detailed 3D information of the forest

upper canopy; however it provides very limited information of the forest understory structure

and mass. Some experiments have proven that TLS can be used to provide high detail

information on the understory with a limited spatial coverage due to the line-of-sight obstruction

caused by the same trees. An opportunity was identified for which dataset from both airborne

and terrestrial platforms could be merged. The tested approach consisted of developing

techniques for geo-referencing TLS data set to achieve a seamless fusion with ALM data to

generate high density point cloud of forest plots from which forestry metrics can be derived. This

application is fully described in Chapter 7.









CHAPTER 5
ST. AUGUSTINE BEACH EROSION HOT SPOT MAPPING

5.1 Motivation

Costal engineers and scientists have known that beaches are subject to both natural or

artificially induced sediment transport. With the execution of large beach fills projects along the

coasts of America in the 1980's and early 1990's, valuable experience was gained in long-term

maintenance and beach-monitoring programs. That experience led to the recognition, systematic

monitoring and study of Erosion Hot Spots or EHS. An EHS is an area that erodes more rapidly

than the adjacent beaches or more rapidly than anticipated during beach fill design. Today

knowledge of coastal processes is capable of explaining what causes most types of EHSs and to

formulate appropriate correction actions. EHSs can be classified and defined by several metrics

such as loss of beach width (recession rate), loss of sediment volume (erosion rate), percentage

of fill remaining of the amount placed, and perception of how a fill should perform relative to

adjacent beaches or to historic rate (Kraus & Galgano, 2001).

Airborne Laser Mapping technology has been extensively used to study large scale beach

erosion. ALSM data covers a long stretch of beach with a moderate sample density of

approximately 1 laser return per square meter (however, most current ALSM systems such as the

Optech Gemini are capable of high pulse rates >100 kHz, with these type of systems 8 to 10 laser

returns per square meter can be achieved). This sampling capability enables the detection of

submeter-scale changes in shoreline position and dune heights over periods of a few months.

However, it might not be as effective for mapping short term, small scale variations that are

characteristic of some localized erosion hot spots. The M-TLSS, on the other hand, can provide

high density point clouds (centimeter scale point spacing) of smaller areas known to be highly









prone to erosion. This chapter will discuss the application of M-TLSS as a complement to ALSM

in the study of beach morphology in the St. Augustine, Florida area.

5.2 Use of LIDAR Technology

Airborne LIDAR has been used since 1996 to study beach erosion. Early projects included

the Airborne LIDAR Assessment of Coastal Erosion (ALACE) that was a partnership between

NOAA, NASA, and the USGS and the Laser Swath-mapping Evaluation and Resurvey (LASER)

undertaken by the University of Florida, the Florida Department of Environmental Protection

and the Florida Department of Transportation. Numerous papers have been published on the

subject proving the success of this application.

At the time of this writing, a literature search for the application of terrestrial laser scanners

to study erosion hot spots yielded no results. The only reference was to a poster presentation at

the 2005 meeting of The Geological Society of America (GSA). The abstract describes the use of

a Terrestrial Laser Scanner to map a beach re-nourishment plan covering 8.59 km of shoreline at

Folly Beach, South Carolina (Kaufman et al. 2005).

5.3 Data Collection

A known erosion hot spot (EHS) along the St. Augustine Beach, Fl. area was selected for

this study. The EHS is located on the beach in front of the St. Augustine Beach Front Resort (300

Ala Beach Blvd, St. Augustine, Fl.). Figure 5-1 shows a map and a near infrared aerial photo of

the study site, the orange polygon defines the mapped area. Data were collected on 4 dates, one

prior the beginning of the Hurricane season on May 23, and 3 takes at two week intervals on

October 28, November 10 and November 25, 2006. GPS data for the geo-referencing of the data

set were also collected on the first three takes using Astech Z-Extreme and Astech Z-Surveyor

geodetic grade receivers.










1 A -. 2













Figure 5-1. Beach erosion hot spot study site location. A) Florida map, orange circle marks the
location of St. Augustine. B) Aerial infrared photograph of the study site, orange
polygon marks the specific mapped area.

5.4 Data Processing

The first step in the preliminary processing was the merging of individual scans taken for

each day into a single point cloud in the sensor based coordinate system (XYZ). Polywork's

Inspect N-pair common point method was used for merging the scans. The next step was the

geo-referencing of the point clouds; on the first take 7 GPS control stations were deployed, for

these stations both XYZ and Easting, Northing, and Height coordinates are available which

allows for the computation of a 3D solid rotation and translation transformation. Figure 5-2

shows a rendering of the March 23rd geo-referenced point cloud, NAD83 was used as horizontal

datum and NAVD88 as vertical datum. In Table 5-1 the control points coordinates used for the

geo-referencing transformation and the transformation residuals for the first data set are

presented. To achieve comparable datasets the geo-referencing of the last 3 point clouds were

performed using 12 common points to the first geo-referenced point cloud. The RMS from the

process of coregistration of the point clouds based on the twelve common points are presented in

Table 5-2.

































gure 5-2. Rendering of the March 23', 2006 dataset.


A set of control features that were used to visually verify the coregistration of the point

clouds are presented in Figure 5-3. Different colors are used to identify point clouds collected on

different days. From this figure it can be verified the relatively good agreement in coregistration

among the scans as described by the RMS values of Table 5-2.

Table 5-1. Control points used for the geo-referencing of the March 23, 2006 dataset.
Sensor Coordinates [m] Geodetic Coordinates [m] Transformation Residuals
X Y Z Easting Northing E Hgt Est Nrth Hgt
P1 -27.497 -46.728 -0.623 474315.01 3303135.38 -23.59 -0.083 0.034 0.007
P2 -78.630 -32.978 -1.623 474346.58 3303177.83 -24.24 0.053 -0.002 -0.022
P3 5.096 63.037 -2.603 474404.80 3303064.44 -24.42 0.054 0.064 0.077
P4 58.391 108.561 1.111 474427.29 3302998.09 -20.61 0.088 0.067 0.366
P5 -166.835 -38.306 -1.419 474374.32 3303261.80 -23.89 -0.001 -0.066 0.039
P6 0.000 0.000 0.000 474348.44 3303092.52 -22.23 -0.284 0.050 -0.249
P7 32.780 171.860 0.959 474495.48 3302998.67 -19.34 0.174 -0.148 -0.218
RMS of Transformation Residuals [m] 0.137 0.074 0.190


Table 5-2. Point clouds coregistration RMS values.

Point Clouds Coregistration RMS [m]
Easting Northing Height
1-2 0.108 0.237 0.135
1-3 0.090 0.211 0.116
1-4 0.086 0.133 0.098



























Figure 5-3. Features used to check the co registration of the point clouds.

The third step consisted of cropping the point clouds to the specific area of interest as


defined by the orange polygon on Figure 5-1 B and filtering to remove all the non-surface


objects (people, beach chairs, etc.). The fourth and final step was the creation of 10 cm spacing


regular grids by the method of triangulation with linear interpolation using Surfer & Matlab


software packages. Figure 5-4 present image maps created from the 10 cm elevation grids.


10 cm Grids generated from M-TLSS data of the St. Augustine Pier Area.
Datum: NAD_83. Coordinates: UTM Zone 17 in meters. Real Northing = Ploted Northing + 3,000,000. Vertical Component: Ellipsoidal Height.
May 23, 2006 Oct28, 2006 Nov 10, 2006 Nov 25, 2006

30523r 303230 303230 303230

3 210 303210 303210 303210

305'O 3001 303190 303190 303190

3031:0 303170 303170 303170 ;

30:~. 1 1 303150 303150 303150

3031 X. 303130 303130 303130 I '

3031 10 303110 303110 303110

303090 303090 303090 303090 I *

303070 303070 303070 303070 .

03050 1 303050 303050 303050

_0r.0 I o303030 303030 303030

ire 5 I e m frm te 1 c e tin ri


Figure 5-4. Image maps from the 10 cm elevation grids.










5.5 Results


5.5.1 Elevation Changes

Elevation changes were computed for both short and long term periods. Using Surfer the

elevation grids were difference two at a time. Two comparisons were performed for long term

change (May 23 to October 28 and May 23 to November 25) and two for short term change

(October 28 to November 10 and November 10 to November 25). The results of the elevation

change detection are presented as difference grids on Figure 5-5. The long term change grids

reflect an average of 20 cm of reduction in the berm elevation and a 1.4 to 2 m difference in

elevation between the berm and the surf zone. The short term change grids show a general

preservation of the berm elevation and an average difference of Im in the berm to surf zone

elevation.

Elevation Change Computed from 10 cm Grids of the St. Augustine Pier Area.
Datum: NAD_83. Coordinates: UTM Zone 17 in meters. Real Northing = Ploted Northing + 3,000,000. Vertical Component: Delta Ellipsoidal Height.
Long Term Monitoring Short Term Monitoring
Oct 28 May 23 Nov 25 May 23 Nov 10 Oct 28 Nov 25 Nov 10



Figure 5-5. Image maps from the elevation change grids.1.4



0.4




41.2
3D3110 -1.4

--,8
-.2
1 303070.4




474355 474375 474395 474415 474435 474355 474375 4745 474415 474435 474355 474375 474395 474415 474435 474355 47475 474395 474415 474435
Figure 5-5. Image maps from the elevation change grids.









5.5.2 Volume Changes

Total volume differences were computed and normalized to obtain lost volumes per unit

beach length and rates of lost volume per unit beach length. For planners and engineers the total

volume lost is important because that translates directly to the renourishment costs, however

absolute measurements are difficult to use for comparisons; lost volumes per unit beach length

and rates of lost volume per unit beach length to provide a better understanding of the magnitude

of the change. Table 5-3 summarizes some of the volume computations performed from the

elevation grids.

Table 5-3. Summary of volume change computations.
Volume Volume
Lost Volume loss Volue
loss/beach unit loss/beach unit
From To Days volume rate length length rate
length length rate
[m3] [m3/day] [m3/m] [m3/m day]
5/23/2006 11/25/2006 186 13168.245 70.79702 59.31642 0.318905
5/23/2006 10/28/2006 158 9166.72 58.01722 41.29153 0.261339
10/28/2006 11/25/2006 28 4945.707 176.6324 22.27796 0.795641
10/28/2006 11/10/2006 13 5535.53 425.81 24.93482 1.918063

5.5.3 Beach Line and Crest of Berm Extraction From the Grids

The recession of the beach line is an important phenomenon to record and quantify. The

beach line is defined by the mean higher high water (MHHW) line; which is an average of the

higher high water height of each tidal day over nearly 19 years. For the study area the mean

higher high water line is determined at 0.6 meters for the NAVD88 vertical datum. Because the

elevation grids were created using ellipsoids heights a conversion from orthometric height to

ellipsoidal height was performed. The average Geoid separation was found to be 28.612 meters

so the beach line was extracted from the -28.012 meters ellipsoidal height contour. Similarly the

crest of berm was extracted from the -26.5 meters contour, Figure 5-6 contains plots with the

beach and crest ofberm lines.










Extracted Beach Lne Geoid b=0 6 m NAVD88 Eaxracted Crest of Berm Postion Geoid H=2 1 m NAVD88

















4I I. I I






Figure 5-6. Beach line and crest of berm position plots for each of the data collection dates.

5.5.4 Across Beach Profile Extraction

The traditional data collection method for studying beach erosion is transect sampling.

FDEP standard require cross shore transects at every 1,000 feet apart along shore; with a

collection interval not to exceed 25 feet and at all grade breaks. From the M-TLS dataset

generated grids, transects can be obtained automatically at higher resolutions such as presented

in Figure 5-7. To do any kind of interpretation with this transects is necessary to specify a

tolerance in the horizontal and vertical components due to the accuracy of the coregistration

procedure. This tolerance is in the same order of the highest residual of the coregistration control

points. These residuals were in the order of 11cm in the East-West direction, 24 cm in the North-

South direction and 13 cm in the vertical dimension. From these transects it can be seen that the

beach line and the berm line receded twenty meters on average. However, the most interesting














result is that half of that recession occurred in the last four week period of the complete 27 week



observation program. This accelerated erosion can be related to the appearance of the North



Eastern winds which are recognized as the main sediment transport mechanism of that area. The



berm maintained a constant height of roughly 3 meters above the beach line.


Extracted Across Beach Profile, Northing 3,303,114 5 meters


UTM Zone 17 Easting [meters) in


Extracted Across Beach Profile, Northing 3,303,172.0 meters
May 23
O- ct28 --
Nov10
- - - -- N ov25 ---

--- --- -- T ......... 4744 4



i.. . . . . .. .. I .. .. .. .. -- - -- - .. . . .


35 4.7436 47437 4 7438 4.7439 4744 47441
UTM Zone 17 Easting [meters]


Extracted Across Beach Profile, Northing 3,303,199 5 meters


4 7442 4 7443


Extracted Across Beach Profile, Northing 3,303,227 0 meters
-2- May 23
25- -------- ------------- ----------- -------------- ----------------- --------------- O 2
Nov 10



-26 ----- ------- ------ T ------ ----- --- ------------------------------ ---------------- -------------- ---------------,---------
-27 ----------------- I --- -- No--25




-29 --

-30 l i i


77435 4 7436 4 7437 4 7438 4 7439 4 744 4 7441 4 7442 4 744


Figure 5-7. Beach profiles extracted from the grids showing the recession of the berm.


47444
x 1n


4 7444
x 105













5.6 Comparison Between Traditional Methods and M-TLS


Based on the resolution established by FDEP standards for beach profile topographic


surveying, simulated GPS or leveling profiles were generated to illustrate the difference in


resolution between traditional surveying methods and these achieved by the M-TLS. The results


are presented in Figure 5-8, these show that the traditional methods do not capture the small


scale details of the beach and berm surface. It can also be seen that the traditional method can


over estimate beach erosion as they do not sample properly the berm crest.


Etractad Acrass Bc ,c Profile, NrtiHng 3,303,114.5 m. Ruy 23











EXtcL5d A Eh Pri Nort 3,33,15 m. Ot28
E -M-5










E-.tiU IMz 17[

EXcracted ACIri Elach Prlle, Nortln 3,3W.114 5- m. Nov10
-5



47430 474O 474370 47 0 474390m 474400 474 0 47












Eating U zne 17 [

ExIctmced Across BEdch Profile, N t-ring 3,303,114.5 m. ctv25
-2 5I -











2 5 |-- ------------------------------------------------------------
-2.- ----- ------ ..............2:.5.:.... .:.^ '*^ ---- ------------------ ----- -----


Etirg LM zoner 17 [n]
S M-Tradtioa'Suvey

Figure 5-8. Comparison of profile resolution generated from traditional methods and M-TLS.







83


I


L
-2


47 47 47 0
474350 474360 474370 474380 474390 474400 474410 4744 33









CHAPTER 6
SOIL ROUGHNESS METRICS DETERMINATION

6.1 Motivation

The application of active and passive microwave remote sensing is increasing in the field

of soil science as a tool to map soil properties. It is possible to use microwave backscattering to

extract geophysical surface parameters such as soil moisture content and soil surface roughness.

The scattering of microwave energy is determined by the sensor parameters such as wavelength,

polarization and observation geometry; it also depends on the surface dielectric properties and

roughness. In theory it is a simple computation to solve for one of the parameters (moisture or

roughness) having prior knowledge of the other and of the microwave energy backscattering.

However many practical problem arise when trying to parameterize the soil surface roughness.

(Callens et al.; 2006 & Zribi et al.; 2000) Traditionally soil roughness has been characterized as a

single scale process obtained from 2D profiles and parameterized by the root mean square

(RMS) height (s), the correlation length (1) and autocorrelation function (p(e)). The main

limitations of this modeling as reported in the literature are:

* The soil surface and its roughness are multiscale in nature, simplifying them to single scale
parameters implies a loss of information.

* Theoretical and field data have shown that different values of the roughness parameters,
especially on "s" and "C', can be obtained from the same surface as a function of the profile
length, discretization interval, the instrument resolution, and the overall shape of the
profile.

* The surfaces are assumed to follow Exponential or Gaussian distributions without having
the ability to check these assumptions.

To overcome these limitations and the inadequacy of the single scale models in describing

complex soil surfaces, several alternative multiscale roughness description models have been

proposed. They include the mixture of small and large single scale features, the use of random

fractals and fractal dimensions to describe the surface (Davidson et al., 2000), and the use of









plane facets and 3D statistical analysis (Zribi et al., 2000). However, the universally accepted

theoretical microwave backscattering models such as the Small Perturbation Model (SPM),

Kirchoff Approximation (KA) and the Integral Equation Model (IEM) continue to require only

the single scale parameters as inputs.

A novel approach using the M-TLS to scan scattering surfaces to generate 3D terrain

models was explored. From the 3D models the distribution function of the single scale

parameters can be obtained, which is expected to fully describe the surface roughness, thus

overcoming the limitation of under representation produced by the profiling sampling methods.

6.2 Use of LIDAR Technology

The traditional methods of obtaining the soil profiles to compute the roughness parameters

by mechanical means are the Needle-like Profiler and the Mesh Board. A complete description

of these mechanical profilers and the data collection procedures is provided by Mattia et al.

(2003). The main disadvantage of these mechanical methods is that they tend to disturb the

surface that is under study. Mesh Boards have to be hammered into the soil, while Needle-like

Profilers tend to penetrate into the surface yielding noisy measurements of the heights. Modern

methods are aimed at not disturbing the surface; these methods use non-contact instruments such

as laser profilers optical imagers and acoustic backscatter instruments (Mattia et al., 2003; Zribi

et al., 2000; Oelze et al., 2003).

Laser profilers have been used intensely for soil roughness digitizing; most of them are

commercial systems capable of high spatial resolutions of the order of 1 mm or less but limited

to relatively short profile lengths, usually no more than a few meters. Special laser profilers have

been develop to allow the measurement of longer profiles, such as the CESBIO-ESA laser

profiler, which is capable of acquiring roughness profiles up to 25 m long (Davidson et al.,

2000).









Whether using mechanical or non-contact techniques, there are sources of error that are

common when characterizing the soil roughness. These are the truncation Error, which arises

from measuring relatively short profiles; and profiler error, which is due to the intrinsic

limitations of a measurement method (Mattia et al., 2003). It is expected that by full surface

digitizing, using the M-TLS, the errors from these sources will be drastically reduced. A search

of the literature found no journal papers on the use of terrestrial laser scanner to digitize a surface

and later compute its roughness parameters.

6.3 Data Collection

Data was collected using the ILRIS in an experimental plot at the University of Florida

Plant Science Research and Education Center at Citra, Florida. The collected data contains the

soil surface of the footprints of two passive microwave radiometers operating in the L and C

bands. Three different time samples were collected, the first just after soil tilling, the second after

corn planting and the third after the crop was harvested. An additional data set that contains a

larger horizontal variation was collected on a commercial plantation near Hastings, Florida. A

description of the collected data is provided in Table 6-1.

Table 6-1. Soil roughness collected datasets
Dataset ID Collection Date Conditions Radiometer Footprint
Field
Field March 08, 2006 After Tilling Mh B
Field3 Mesh Board Test
Citra01 L
Citra02 C
Citra03 Mesh Board Test
Citra04 March 10, 2006 After Planting Mesh Board Test
Citra04 Mesh Board Test
Citra05 L
Citra06 C
CitraCBand C
CitraLBand May 30, 2006 After Harvesting L
CitiaLBand L
Hastings October 28, 2006 Plowed and Planted N/A









Two traditional meshboard measurements were made near the L Band and C Band

radiometer footprints with the soil tilled. For the present analysis only two data sets will be

considered: Citra 02 and Hastings.

6.4 Data Processing

The first step in the process was to prepare the point clouds for analysis. For each test area

this consisted of converting the existing tilted point cloud terrain into that of a flat level terrain.

The angles required to perform the transformation were determined by fitting a plane to selected

ground points, finding the normal vector and computing the rotation angles about the X and Y

axes. The rotation was performed using TerraScan transformation module. Once leveled the

point cloud was rotated about the Z axis to align the radiometer footprint axes to the point cloud

X and Y axes. Finally the point cloud was cropped to a 4 x 6 meter plot corresponding to the

radiometer footprint. Figure 6-1 shows renderings of the raw and a rectified and point cloud.


gure o-i. "ataset preprocessing steps. Kenaering or A) raw point clouas ana t) recunea point
cloud from the Citra 02.









From the cropped and rectified point clouds regular grids with one cm cell spacing were

created using a triangulation with linear interpolation gridding function in Matlab. Figure 6-2

shows image maps of the elevation grids for two of the analyzed dataset.

A UB rTini eight








n i

















Figure 6-2. Renderings of the 1 cm elevation grids. A) Citra 02. B) Hastings.
i ii i i



40 I
Ulli.


t .5 pro 1 2 25 3 plot were comp
XIlalarlI .5 1 1.5 2 25 3

Figure 6-2. Renderings of the 1 cm elevaation grids. A) Citra 02. B)Hastings.

The cropped point clouds and the grids were used to derive the roughness metrics. The

expressions for the height RMS, correlation length and autocorrelation function for 2D profiles

are given in Table 6-2. Several tests were performed to prove the advantages of using MTLS

datasets to extract the single scale metrics over the traditional profiling methods. These tests are:

* From the regular grid a set of random 2D profiles was extracted parallel to the X and Y axes.
The root mean square height (s) and the correlation length (1) for each profile were computed
using the traditional formulas. Mean and standard deviations of the roughness parameters of
the profiles of the same plot were computed.

* The formulas used to compute root mean square height (s) and correlation length (1) from the
2D transects were extended to 3D surfaces and are used to compute the roughness metrics for









the two plots. The correlation was also extended from ID to 2D, so the correlation length
could be found for either X and Y, or a combination of both directions. The 2D correlation
length was converted into a ID length and distribution functions of correlation lengths were
obtained.

* Comparisons were made between the averaged roughness parameters values from the
random profiles with the ones obtained from the 3D models.

* Test were performed to verify the assumption that the autocorrelation functions follow
Exponential or Gaussian forms.

Table 6-2 Definition of soil roughness parameters.
Callens et al. Thoma et al.
Height mean 1 N
NZ
Height RMS I1 S 1 2-
s= zI 2 -Nx2 s= /-(z,-)2
N-l N ,
Normalized z) [(z- Xz- )]
correlation V) pX+, 1
function p(h)= p(jA)= p(h)= p(jAx)= =1
function (-Zz f)2
1= 1 l
Correlation length such that )
1 such that : p(1) = -
e
Exponential lh
autocorrelation p(h /
function p)= e
Gaussian
autocorrelation p(h) e
function pV)=

6.5 Results

6.5.1 Simulated Profiling Results

From the elevation grids random transects were extracted parallel to the X & Y axes to

simulate traditional soil roughness profiling techniques. Using the 2D formulas the roughness

metrics were computed to verify the variability of the metrics with the profile selection. Figures

6-4 to 6-7 show the extracted profiles and their respective normalized autocorrelation plot from

the Citra 02 and Hasting datasets. The red line in the autocorrelation plots represents the 1/e

value used to determine the correlation length.














A D.rt




> -l ld


Grid Entyacted Pmrils Prail tD Ihe X Axi5
I I I" : -- Pr 1 -Il
; ,. .." -- Prfa 2Y=m


... .- .... .. .


I i I i i
as I .s t 2 2.i 3 IS
X[Mltars|
Nrmlized Atccolation


-0 -300 -2 -100 0 100 B3 30M 41
Lugs IG| d unts]
Figure 6-4. Roughness parameter plots for the Citra 02 dataset parallel to the X axis. A)

Extracted profiles. B) Normalized autocorrelation plot.


A Grid Eitrxted Priles- Parala to t Y A ,;
I I I


0.,L ,- *-. .. ... ... ; ....... ........ .. ..... .. ... .. P.. ..... ... .
S. ., = -. ., ...J. *, . -.

o :.i . .. ..


0


0 1 2 3 4 5 6 7
X IMpi BJ
Normalized Autarrelatian



--- ---- -~" 3 )-.2 75m." i
l--



.G0-11 a1 2 J 4W


Lar [Gnd0m1Ij n

Figure 6-5. Roughness parameter plots for the Citra 02 dataset parallel to the Y axis. A)

Extracted profiles. B) Normalized autocorrelation plot.


A Gdid Enirated Pitdls Parael I IhLX i is
a. -


a
S I.


I. I.N
~~-~`. ---~~~--'~.~e .... ., ~ ,. : .~


SIMeltsl


- P roN 1 Y=li
-- ProtneY=.-t


B 1



~ cs
;r
I
a.
r
_I


l -3C -200 -100 a 100 200 3
LaIg Grid ufits|

Figure 6-6. Roughness parameter plots for the Hastings dataset parallel to the X axis. A)

Extracted profiles. B) Normalized autocorrelation plot.


I


;-- -- -- --


i-/~t~ i I~-~
- /


c-


~s~- ';-4
~--1.










A


Grid E+t.-!d; Fauis PuigIl D3 IN AMip


F-I- 1 XJ:Sm

.. -.. ........... .. ... .
-K-.*7* _-..





3 NcmLhA Pd a flciIJ
: CY1T----------- -- ------ | ---- ^---------- --








Figure 6-7. Roughness parameter plots for the Hastings dataset parallel to the Y axis. A)
Extracted profiles. B) Normalized autocorrelation plot.

The results of the roughness metrics are summarized in Table 6-3 for the Citra 02 data set

and in Table 6-4 for Hastings. It can be observed from the tables that the mean height RMS from

profiles in both directions (parallel to Y axis and parallel to X axis) are relatively close (3 to 25%

variation), however the correlation length can vary greatly between directions (39-77%).

Table 6-3. Soil roughness parameters results from random profiles for the Citra 02 grid.
Data set RMS regular spacing [m] 1/e correlation length [m]
Citra 02 Y=1 m 0.0133 2.1718
Citra 02 Y=3 m 0.0100 2.1657
Citra 02 Y=5 m 0.0108 2.3727
Citra 02 X=0.75 m 0.0115 4.0464
Citra 02 X=1.75 m 0.0079 3.8960
Citra 02 X=2.75 m 0.0077 3.9844
Mean parallel to X 0.0114 2.2367
Mean parallel to Y 0.0090 3.9756
Overall mean 0.0102 3.1062
Overall cy 0.0022 0.9565

Table 6-4. Soil roughness parameters results from random profiles for the Hastings grid.
Data Set RMS Regular Spacing [m] 1/e Correlation Length [m]
Hastings Y= m 0.0277 1.9087
Hastings Y=3 m 0.0543 2.0181
Hastings Y=5 m 0.0715 2.0127
Hastings X=1 m 0.0458 2.6581
Hastings X=3 m 0.0497 2.9350
Hastings X=5 m 0.0633 2.6739
Mean parallel to X 0.051 1.980
Mean parallel to Y 0.053 2.756
Overall Mean 0.052 2.368









6.5.2 Extension of the 2D Formulas for a 3D Surface.

The height RMS formula provided for 2D profiles in Table 6-2 is valid without

modification for the 3D soil surface. For 2D profiles the correlation length is defined as the

distance for which the value of the normalized ID autocorrelation is 1/e. The concept can be

extended for a 3D surface by computing a 2D autocorrelation with two parameters (x lag and y

lag) which results in a 3D surface. The 1/e contour can be located on that surface, and the

correlation length can be defined as the scaled distance (considering grid element size) from the

origin to a particular point on the contour. The advantage of this method over the traditional

profiling method is that from this a complete distribution function of the correlation length is

obtained rather than a single value.

A ,,tu B C











........e 6 .- Citra n r aie h h t r/ m C no p' l
F"ig
W IIAX 7
Fiue68 I A lo an l



04. U (

I I










Figure 6-8. Citra normalized height autocorrelation. A) 3D plot. B) Color map. C) Contour plot.




















'f
V

A.


i
141
II, .,
di~l
161
ni.;
GI ,


,1.
n
y
icn
~
ar
i
II
-rm


t -I**P.~ I


~w.





a- i
a .j II;


I...I II
II'

~ -~ f 2

i


,t .


Figure 6-9. Hastings height autocorrelation. A) 3D plot. B) Color map. C) Contour plot.


*I,-


k4LI


.a48
~li Crn


t _




____Hj~


0 SE


Cieladiin LmngKh Hsii gnu


400-

300 -
200.
100
0.


tN4 ~ CD ~~ ~N
IiP M


Figure 6-10. Correlation lengths extraction for the Citra 02 dataset. A) 1/e contour plot of the 2D
autocorrelation function. B) Unfolded correlation length plot. C) Distribution function
of correlation lengths.


~srrrrr~n~h


: r
~m ~m


v oCw"arMranL


2w X(O


Colkur spleu
Xw M~Q
C~Swpb











V. C2 .r BP


E C
0
cn


Con*Mcn Um Legfi 1tcgrwn


300.f
250-

150 LU

50-


Figure 6-11. Correlation lengths extraction for the Hastings dataset. A) 1/e contour plot of the 2D
autocorrelation function. B) Unfolded correlation length plot. C) Distribution function
of correlation lengths.

Graphs of the 2D autocorrelation are presented in figures 6.8 and 6.9, part A of the figures

are 3D plots that show the normalized autocorrelation function as a function of the X and Y lag

units. Part B is an image map of the same normalized autocorrelation values and part C is a

contour plot of the correlation values. From the normalized autocorrelation the 1/e contour can

be extracted as shown in figures 6.10 and 6.11 part A. From each point in the contour its

direction and correlation length can be computed. In part B of figures 6.10 and 6.11 the contour

is unfolded and the correlation length for each point is computed and plotted without taking into

consideration the direction. The obtained correlation length are then grouped and binned to

create the histograms on part C of the figures. Table 6-5 provides a summary of the roughness

parameters values extracted from the 3D surface models. From this table and figures 6.10C and









6.11C it can be seen that the mean correlation length is not necessarily a good descriptor of the

distribution.

Table 6-5. Soil roughness parameters from 3D surface models
Roughness Parameter Citra 02 Hastings
(s) RMS height from point cloud [m] 0.013546 0.05013
(s) RMS height from grid model [m] 0.012674 0.0547
(A) Mean correlation length [grid units] 271.78 215.25
(t) Mean correlation length [m] 2.7178 2.1525
Standard deviation of correlation length [grid units] 60.42 52.23
Standard deviation of correlation length [m] 0.6042 0.5223

6.5.3 Comparisons of Roughness Metrics From Profiles vs. Full Surface

The values of the roughness parameters obtained from the random profiles and the 3D

surface are tabulated for comparison in tables 6.6 and 6.7.

Table 6-6. Comparison of soil roughness parameters for Citra 02 from 3D surface models and
random generated profiles.
Transects// Transects// All
Roughness Parameter 3D Surface T
to X to Y transects
(s) RMS height [m] 0.012674 0.0114 0.0090 0.0102
(1) mean correlation length [m] 2.7178 2.2367 3.9756 3.1062
Min correlation length [m] 2.0597
Max correlation length [m] 3.9651
Mode correlation length [m] 2.2
y of correlation length [m] 0.6042 0.1178 0.0756 0.9565

Table 6-7. Comparison of soil roughness parameters for Hastings 02 from 3D surface models
and random generated profiles.
Transects// Transects// All
Roughness Parameter 3D Surface T
to X to Y transects
(s) RMS height [m] 0.0547 0.051 0.053 0.052
(1) mean correlation length [m] 2.1525 1.980 2.756 2.368
Min correlation length [m] 1.4200
Max correlation length [m] 3.1620
Mode correlation length [m] 2.1
y of correlation length [m] 0.5223 0.062 0.156 0.438

6.5.4 Distribution Functions of the 3D Correlations Lengths.

From Table 6-2 it can be seen that the distributions of the correlation lengths are assumed

to follow either and exponential autocorrelation function (ACF) or a Gaussian autocorrelation

function form. One of the advantages of the full surface digitizing is that the complete









distribution function of the correlation lengths can be obtained. Figure 6-12 shows the

distribution and compares it with a normal Gaussian distribution. It can be clearly seen that the

distribution is not close to either an exponential or Gaussian distributions.

A B
careWon Lenglh Hilelgnamin a02 Dliaset ConTeleSon Lgh HMogam HBain Qtaet





U 20 0,2 100- 0,2
100 .1 I 0
n M 10, 2006 m o0 nt w n f in f


LtPng.hln Lengt [Fri
Bper r or Data- !nal N !Cftkthn =l Exprtim Dta-N=Varne :it9MtLcn

Figure 6-12. Comparison of experimental correlation length distributions with respect to the
assumed normal distribution. A) for the Citra 02 dataset B) for the Hastings dataset.

6.6 Comparison with the Traditional Meshboard Method.

On March 10, 2006 meshboard measurements were made near the footprints of the

radiometers. Figure 6-13 illustrate the meshboard technique, a picture is taken of the interface

between the soil and the board. Image processing software is used to digitize the coordinates of

the soil surface based on the mesh printed on the board.


Figure 6-13. Meshboard used to digitize










Figure 6-14 show the digitized soil surface transect from one of the footprint areas based

on the meshboard method. On part A the irregular spaced points are plotted, part B plots a

regular spacing (1 cm) sampling interpolation. Part C is a plot of the normalized 1D

autocorrelation.

A Irregular Spaced Terrain Profile
3 .......... :......... J......... J.......... ........... I ......... ........... ......... ..................... L.

....... I .V L ,- % ....-..
; ... ......... .... .. ........
---------- -------- ---- --- --- ----- ----- ---- ------- --------- ----------
0 0 40 60 80 100 120 140 160 180 200
X [cm]
B Regular Spaced Terrain Profile




0 2 40 60 8) 100 120 40 160 160 211
c X [crn]
Height AutoCorrelation
C ^ -r -- -- --* ** 1 V --- *" I- -- -- -I---------- *-- *--------T ---------- -- --- ---* -- -- ,--- ....-- -----
2- .--- ---------- -.---- t -----.- -------- -------
1 ---- -- -- 1- a ,,

I ------ ------ -- ----- ------
., -, r........ .... ..... ... ...... ... .. .. .. ..... .. .. ........... ............. ..
0 20 40 60 80 100 120 140 160 180 200
F X [cm]
.Height Autocorrelation


------ I -- --- --- -- -I---- ---;-- -----f-- ---- 4 ---- ---- ---4------- 4 --.-----.----. I--'


z -200 -150 -100 -50 0 50 100 150 200
X Lags [Grid uniTrs
Figure 6-14. Plots of meshboard derived data. A) irregular sampled profile B) regular spacing
profile C) normalized height autocorrelation.

The results obtained from the traditional meshboard method are compared with the results

obtained thru the 3D datasets in Table 6-8. There is a disagreement in the values obtained from

the different methods, being the greatest difference the autocorrelation length on the C-band

radiometer footprint.

Table 6-8 Comparison of soil roughness metrics obtained from the traditional and alternative
method.
Data Set RMS all points RMS regular spacing 1/e correlation length
MB Near L Band FPT 0.009158 m 0.009291 m 1.427 m
3D L Band FTP 0.02179 m 0.022513 m 2.2512 m
2. Near C Band FPT 0.013117m 0.012122 m 0.284 m
3D C Band FTP 0.013546 m 0.012674 m 2.7244 m









6.7 Conclusions

Having a 3D data set of a scattering surface allowed us to demonstrate the common sense

knowledge that the values of the derived soil roughness parameters are highly variable

depending on the profile selection. It was also possible to derive the traditional 2D profile

metrics from an entire 3D surface, having not a single value for correlation length but a range of

values that fully describe the scattering surface. This proved the advantage of a full surface

modeling over the traditional profiling methods.









CHAPTER 7
FORESTRY METRICS APPLICATIONS

7.1 Motivation

The estimation of forest structure and volume has for decades been of great interest to the

scientific community because of its ecological and economical importance. Traditional methods

of performing these measurements and estimations, even over a small plot of forest, are labor

intensive and require many hours of field work. Several non-invasive remote sensing

technologies have been tried to simplify these tasks and collect more accurate measurements in

less time.

Over the past several years a great amount of experience has been obtained in the

application of Airborne Laser Mapping (ALM) to study forest structure and estimating its

biomass. ALM provides a large spatial coverage with very detailed 3D information of the forest

upper canopy; however, it provides very limited information of the forest understory structure

and mass. On the other hand, Ground-Based Scanning Laser or Terrestrial Laser Mapping

(TLM) can provide very detailed 3D information on the understory structure with limited spatial

coverage due to the line-of-sight obstruction caused by the trees. In contrast to ALM, which is a

relatively mature technology with commonly accepted data collecting techniques and

procedures, there are no guidelines or uniformly accepted procedures on how to set up a TLM

system to obtain data for forestry applications.

Bibliographical research indicates that very few efforts have been made to use TLM to

investigate Forest Metrics and structure. Previous work has been limited to assess the potential of

TLM for replacing traditional field techniques to determine basic tree & plot metrics such as tree

height, stem diameter at breast height (DBH), stem location and stem density; or as ground

validation for airborne remote sensing technologies.









The long term goal of this project is to develop techniques for a seamless merging of ALM

and TLS observations to generate high density point clouds of forest plots. The final goal is to

develop high resolution laser tomography through a multi-platform and multi-imaging-geometry

data integration to generate a virtual "Forest Cube" for which all metrics can be derived with

high levels of accuracy and precision. With the derived metrics of several of these "Forrest

Cubes" of statistically sampled plots combined with the large spatial coverage of ALM, forest-

wide metrics can be estimated with a high degree of confidence. The first step and shorter term

goal is to develop the required field techniques and procedures for TLM forestry data acquisition

and geo-referencing for a typical small size test plot.

7.2 Use of LIDAR Technology

In the early 1980s LIDAR scanners were developed to obtain high accurate

measurements of surface elevations from airborne and spaceborne platforms. These first systems

are classified as large footprint LIDARs because the intersection of the laser beam with the

surface is a circular or elliptical spot of several meters in extent. The first LIDAR systems

suitable for vegetation and forestry studies had a footprint of less than a meter and were designed

to record only the first received return (Lefsky et al. 1998). Those systems evolved to present day

airborne laser scanners that generally have small footprint and the capability of recording

multiple returns, as well as their intensities. Detailed descriptions of LIDAR technologies

instrumentation and operation can be found in Wehr & Lohr (1999) and Baltsavias (1999).

Ground-Based Laser Scanning or Terrestrial Laser Scanning (TLS), is a relatively new

technology compared with their flying counterparts. They evolved from the Electronic Distance

Measurement (EDM) devices used in traditional surveying. Terrestrial laser mapping systems

produce very accurate 3D data sets of large surfaces rapidly. Other surveying or

photogrammetric techniques require much longer acquisition times and yield much fewer 3D









points. A ground-based laser scanner provides 4 quantities for each sampled point, three are

positional information in a Cartesian reference frame: the (x,y,z) coordinates, and the fourth

quantity is the magnitude or intensity of the return signal. The sampling resolution is measured in

angular units because linear spacing between points depends on the range. For a 2 milliradian

angular resolution the linear spacing between samples is 1 cm at 10m or 10 cm at 100 meters.

The key features of a ground-based laser scanning system are: a) range accuracy: which is

dependent on the pulse duration, b) maximum range: which depends on the laser output power

and the receiver sensitivity, c) scan rate: depends on how fast are the laser and receiving

electronics, d) the angular resolution, e) field of view: the last two depend on the mechanics of

the scanning mirror system, and the last but most important is the f) laser, which is characterized

by its wavelength and beam divergence. More detailed information on TLS can be found in

Lichti et al. (2002).

7.3 Previous Works

Over the last two decades multiple papers on Airborne LIDAR applications to vegetation

and forest studies have been published, discussing the capabilities and limitations of the different

systems and applications. Some of the most complete and interesting papers are: Lefsky, et al.

(1998); Lefsky et al. (2002) and Nasset & Gobakken (2005). The overall conclusion is that

airborne LIDAR mapping is a very powerful tool for deriving and modeling forest metrics;

however, it has the limitation of the low sampling of the understory canopy.

Previous works on the application of TLS to forestry have aimed to compare tree & plot

metrics such as tree height, stem diameter at breast height (dbh), stem location and stem density

derived from TLS with the same metrics derived with traditional ground techniques. These

works have had different approaches with respect to scanning geometry and the shape of the

defined plot. Watt & Donoghue defined circular plots of 0.02 hectare (ha) and obtained a quasi-









complete scan of the area with only two scans from opposite points. Hopkinson et al. defined 35

m x 35m (0.12 ha) square plots and performed 5 scans per plot. Both studies concluded that TLS

was a powerful tool in determining forest variables; however it is limited by obstructions of the

line of sight that determines the useful range of this type of instrument.

The above papers prove the usefulness of both airborne and terrestrial based systems in

the estimation of forest metrics. The work presented here is aimed at developing data acquisition

and processing techniques that will enable the combination of data from multiple platforms (i.e.

Airborne & Terrestrial) and geometries that will produce a complete well sampled data set over

the complete structure of a small plot of the forest.

7.4 Data Collection

The test site for the experiment was at the Intensive Management Practice Assessment

Center (IMPAC) operated by the Forest Biology Research Cooperative (FBRC), and located 10

km north of Gainesville, Florida USA. The center is a research plantation of the southern pine

species loblolly (Pinus taeda L.) and slash (Pinus elliottii var. elliottii). The Airborne Laser data

was collected on October, 2005 using the UF Optech Inc. ALTM 1233 laser mapping system

flown on a Cessna 337. The airborne system, when operating at a flight height of 600 m AGL, at

a flight speed of 60 m/s and a LASER pulse repetition frequency of 33 kHz yields a point density

of roughly 1 laser return per square meter. Two sets of terrestrial Laser Data were collected, the

first in November 2005 and the second in June of 2006 using the UF Mobile Terrestrial Laser

Scanner (M-TLS) system based on the Optech ILRIS-3D terrestrial imaging LIDAR. Figure 7-1

presents an aerial photo of the test site, and Figure 7-2 presents a shaded relief model rendered

from the airborne laser scanner data.

































Figure 7-1. Aerial photographs of the test site. Downloaded from the Florida Department of
Environmental Protection, Land Boundary Information System.
http://data.labins.org/2003/MappingData/DOQQ/doqq.cfm

The terrestrial laser mapping data consisted of five independent scans taken from different

angles and elevations. The total number of return points in the terrestrial clouds was 3,765,084.

The TLM instrument provides a set of points (x,y,z,I). The first three points provide the spatial

information of the scanned surface point. These x,y,z coordinates are distances referred on an

orthogonal frame of reference whose origin is the scanner sensor head. Geo-referencing was

accomplished by setting GPS control points that were included in the scans, and later

determining the geodetic coordinates of these control points. Data was collected using

ASHTECH Z-XTREME receivers, and ASHTECH Choke Ring Antennas Model 700936 Rev D.

A total of 8 stations were surveyed, 6 corresponding to the GPS control points and 2 for the

scanning locations.











3293400


3293200




374400 374600 374800 375000 375200 375400 375600
Figure 7-2. Shaded relief digital elevation model rendered from the airborne laser scanner data of
the test site. The edge effects are caused by the method used to interpolate the LIDAR
observed data to obtain a uniform spaced (gridded) surface model.

7.5 Data Processing

The five individual point clouds were merged into a single data set in a single sensor

coordinate frame using the Innovmetric

(http://www.innovmetric.com/Manufacturing/home.aspx) Polyworks Inspect software employing

the n common point alignment procedure. The GPS reference point (ARP) were used as the

reference points for the alignment. For georeferencing the dataset into a NAD83 Datum

expressed in UTM Zone 17 coordinates, 8 control points were surveyed. Their coordinates in

both geodetic and sensor spaces are listed in the Table 7-1.

Table 7-1. M-TLS data set geo-referencing control network.
Geodetic space coordinates UTM zone 17 Sensor space coordinates
Station Ellipsoid
Easting (X) Northing (Y) Height x z
Height (Z)
GPS1 374997.280 3293231.190 26.067 1.156 21.033 -2.505
GPS2 375004.345 3293235.345 24.982 -6.77 19.81 -4.403
GPS3 375014.622 3293203.487 35.906 -4.275 54.339 -2.766
GPS4 374994.418 3293223.649 25.736 7.06 26.576 -3.955
GPS5 375011.541 3293247.681 25.621 -18.4871 11.63806 -3.71564
GPS6 375011.129 3293242.494 23.451 -15.7089 16.51747 -5.505
IL01 374989.820 3293251.056 23.701 0 0 0
ILO 375015.927 3293268.460 24.739 -30.9398 -5.44228 -2.60267









Following, a 3D translate and rotate conformal transformation was applied using the

Terrasolid (http://www.terrasolid.fi/) Terrascan software. The inputs for the computation of the 9

parameters were the sensor space and geodetic space coordinates of the GPS antennas reference

point (ARP). In Table 7-2 the residuals of the control points based on the geo-referencing

transformation are presented.

Table 7-2. Geo-referencing residuals analysis.
Adjustment Residuals Square of the Residuals
Station Ellipsoid
Easting (X) Northing (Y) Height (Z) x y z
GPS1 -0.0736 0.0591 0.3474 0.005417 0.003493 0.120687
GPS2 -0.0198 0.0011 0.4919 0.000392 1.21E-06 0.241966
GPS3 0.0459 0.2654 -0.3605 0.002107 0.070437 0.12996
GPS4 -0.0541 -0.0475 -0.2618 0.002927 0.002256 0.068539
GPS5 0.0493 -0.0022 0.1346 0.00243 4.84E-06 0.018117
GPS6 This control point was eliminated after a preliminary adjustment
IL01 -0.055 -0.1356 -0.1181 0.003025 0.018387 0.013948
ILO 0.1073 -0.1404 -0.2336 0.011513 0.019712 0.054569
Root Mean Square of the Residuals (Meters) 0.063032 0.127779 0.304205

Once the Terrestrial Point cloud was geo-referenced in the same datum and coordinate

system as the airborne data, both datasets were merged into a single file and were viewed and

rendered. Figure 7-3 is a rendering of the fused point cloud color coded by elevation; the yellow

ellipse indicates the overlaying of the high density terrestrial point cloud. It can be observed the

adequate mapping of the undercanopy structure by the high density of the blue and green color

coded points. Figure 7-4 is a rendering of a 3.0 m wide transect depicting the match between the

airborne and the terrestrial point clouds in the along flightline direction, red dots represent the

low density points obtained from the airborne scanner and the white dots represents the high

density terrestrial point cloud. Figure 7-5 illustrate the match between the airborne and the

terrestrial point clouds in the across flightline direction.
























Figure 7-3. Rendering of the fused point cloud, color coded by elevation.


figure /-4. Kenering ot tusea point cloud cross section in the along the ligntline direction.


section in the cross flightline direction


Figure 7-5. Rendering of fus









Figures 7.6 to 7.8 are renderings of the fused airborne and M-TLS point cloud, some of the

GPS tripods and Antennas that were used as control points can be observed in the renderings.


figure /-o. Kenaerng or tne tusea point cloua, grey scale trom tne laser return intensity.


figure /- /. Rendering ot the tused point cloud, color coded by elevation + laser return intensity.



























figure /-6. Kenaenng or tne top view ot tusea point clouo, color cooae Dy elevation + laser
return intensity

7.6 Results

The objective of merging the airborne and terrestrial point clouds was to synergize the

vantage points of both geometries to develop a dataset that could permit the extraction of the

Tree & Plot Metrics (Tree Height, Stem Diameter at Breast Height DBH, Stem location and

Stem Density) with higher levels of precision and accuracy than those using a single geometry.

From the fused point cloud a "Forest Cube" was extracted, its ground dimensions are 40.4m x

19.16m with an area of 774.15 m2. From this virtual "Cube" all the metrics will be derived.

Figure 7-9 shows renderings of the "Forest Cube".

7.6.1 Stem Density

For stem density determination, a manual count of the standing stems was performed. To

facilitate the count the "cube" was skimmed to 1/20th of its point density, was filtered so that

only points that had heights between 2.3 and 6.3 m above ground level were displayed and was

visualized in an oblique view. Figure 7-10 shows the stem count by row, a total of 96 standing

stems were counted, which yield a stem density of 1 stem per 8.06 m2 or 1240 stems/hectare.





























Figure 7-9. Rendering of the "Forest Cube". A) Color coded by elevation and intensity. B) Color
coded only by elevation.


figure /-1u. Kendenng ot point cloud used tor stem counts.









7.6.2 Stem Location

Previous published works have also performed stem location measurements, however they

were referred to arbitrary sensor coordinates. In this work, the dataset was referenced to geodetic

coordinates, providing absolute coordinates of the stem location. The ideal approach is to fit a

circle to a low cross section of the tree and determine the coordinates of the circle origin and the

circle radius. Polyworks INspect or TerraScan can be used for this purpose.

7.6.3 Stem Diameter at Breast Height DBH

Stem diameter at breast height is defined as the stem diameter at a height of 1.4 m (4.5 ft).

The stem diameter can be determined in conjunction of the stem location. Figure 7-11 illustrates

the fitting of a circle from a section of a point cloud at breast height, from which the stem

locations and DBH were determined.


Figure 7-11. Fitting of a circle at breast height for determining DBH and stem location.









7.6.4 Tree Height

Using Terrascan, points contained in a cylindrical volume centered at the stem location

coordinates were extracted from the "Forest Cube"; the maximum and minimum height points

were obtained and the difference between them was computed yielding the tree height. Figure 7-

12 shows a rendering of the points extracted for a single tree with a height histogram used for the

determination of the tree height.


Figure /-12. Single tree height determination.


~m~rzl~m









7.6.5 Stem Volume

Tree stem volume may be a vague concept; it may be used to refer to the total stem volume

or to the "Merchantable Volume", which is defined for a particular length of at trunk up to which

a particular product may be obtained. There are many ways that foresters estimate the stem

volume, and they vary greatly from place to place, based on the purpose and tree species.

Usually foresters have tables that require tree species, diameter at breast height and the

merchantable height (or total tree height) as inputs. The tables are based on allometric equations

which are empirical regressions that relate stem volume and biomass of species to diameter at

breast height and/or to tree height. For this particular dataset it is impossible to determine the

stem merchantable volume because the trees on the plantation are relatively young and have not

reached the diameter and height at which they are considerable exploitable. However the point

cloud can be used to estimate the total stem volume (including bark) from the ground to the

crown. TerraScan was employed to obtain the tree diameter at different heights as shown in

Figure 7-13.


Figure 7-13. Diameters at different heights for volume computations









Each section between diameter measurements can be approximated as a truncated cone

from which the total stem volume was computed. Table 7-3 summarizes the diameter and height

measurements that were used for the stem volume computation.

Table 7-3. Diameter and heights measurements for stem volume estimation
Height above Segment
Height above Diameter [m] Area [m2] Segment height Sme
ground [m] volume
0 0.2295 0.041367
1.4 0.057914
1.4 0.2295 0.041367
3.747 0.2275 0.040649 2.347 0.096245
5.933 0.1896 0.028234 2.186 0.074878
7.8004 0.1516 0.01805 1.8674 0.042862
9.2853 0.1453 0.016581 1.4849 0.025705
11.0672 0.1485 0.01732 1.7819 0.030202
12.7385 0.1295 0.013171 1.6713 0.025401
15.8157 0.079 0.004902 3.0772 0.02678
18.395 0 0 2.5793 0.004214
Total stem volume [m3] 0.384201
Total stem volume modeled as cone regressed from data [m3] 0.394114637

7.6.6 Tree Biomass Estimation

There are multiple formulas that provide estimates of the total aboveground biomass for all

hardwood and conifer species in the U.S. One formula, presented by Jenkins et al. (2003) is:

bm = e(fl+Aln(dbh))
where:
bm = total aboveground biomass (kg) for trees 2.5-cm dbh and larger

dbh. = diameter at breast height (cm)

Po and pi are fit parameters for each species

For the tree sample that has been worked throughout this chapter the result of the biomass

estimate is:

bm = e(/io+An(dbh)) = e(2 5356+2 43491n(22 95)) = 162.99 [kg]

7.7 Comparison with Traditional Methods

Several of the individual tree metrics were obtained or derived from the traditional forestry

techniques and methods as depicted in Figure 7-14.






















.i64ME"NE -- 10erE- I -2
Figure 7-14. Individual tree metric measurement. A) Circumference at diameter height to
determine DBH. B) Estimation of tree height.

For the tree that was used to derive its metrics from the LIDAR point cloud, 6

measurements of its circumference were measured (62.5, 60.5, 61.2, 61.5, 61.0, 61.8 cm) which

yielded and average DBH of 19.55 cm. Also three estimates of its height were obtained using the

traditional Biltmore Stick (19.28, 17.36, 18.2 m) which yields an average tree height of 18.28 m.

From these measured metrics several others can be derived, a summary of the comparison of the

directly measured and derived metrics obtained from the point cloud and traditional metrics are

presented on Table 7-4.

Table 7-4. Comparison of the metrics derived from traditional methods and from the TLS point
cloud.
Metric Traditional From M-TLS Dier
Metric % Difference
Method point cloud
DBH 0.1955 m 0.2295 m 17.39%
Tree Height 18.28 m 18.395 m 0.63 %
Stem Volume 0.3842 m3
Stem Volume 0.1829 m3 0.2536 m3 7.07%
approximated to Cone
Biomass 110.31 kg 162.99 kg 47.8 %

7.8 Conclusions

It was possible to extract plot and individual tree metrics from the fused data set combining

both airborne and terrestrial scanners. However there is more work to be done in the areas of

merging and geo-referencing of the terrestrial point clouds, such that a more coherent data set









will be available for data extraction. All the work presented here was done in a manual fashion.

However, all of the activities can be written into algorithms that will automate the data extraction

process. The synergy of airborne and terrestrial laser scanning technologies can prove highly

valuable to Forest Science, providing vast amounts of information that can enable the

improvement of the allometric models.









CHAPTER 8
SUMMARY

8.1 Conclusions

Terrestrial Laser Scanners (TLS) are very versatile measuring and mapping tools, but as

with any other tool, to become the "method of choice" they have to prove to be efficient in their

application. This means that they have to be better, faster and cheaper than the alternate

equipment and techniques. Better, faster and cheaper implies both qualitative and quantitative

improvements; such as more measurements with greater accuracy at significant savings in time,

labor and capital investments. In their short lifetime TLS units have proven their advantage over

traditional surveying equipment and methods in many engineering applications such as As-Built

documentation and transportation infrastructure management. With respect to the application of

TLS to scientific mapping, measurement and documentation; it is clear that they are faster and

better than the traditional techniques. What is left is to prove is that they can also be more

economic.

For the tested applications in the fields of geology, paleontology, forestry and coastal

morphology the ILRIS unit and/or the M-TLS system proved to collect data more accurately at

higher resolutions and faster than the conventional techniques. The resulting dataset that can

easily exceed several hundreds of megabytes is both the TLS greatest strength and weakness. On

the positive side the dataset contains a wealth of spatial and spectral information from which an

infinite amount of measurements can be derived, and the researcher can revisit the "virtual site"

over and over again without leaving the office. On the negative side, more data is not necessarily

better. In some cases more data can overwhelm the researcher. It can burry the phenomenon of

interest under millions of unnecessary data points. In some extreme cases the analytical models

used to describe the phenomenon simply can not handle the extreme resolution of a TLS dataset.









Finally, many researchers are interested in obtaining their metrics and can be careless about

performing all the preprocessing (point cloud cropping, merging and geo-referencing) required to

have a usable dataset. However, these disadvantages or weakness can be capitalized into an

opportunity to help develop strong relationships between scientists and geosensing engineers

thru collaborative research projects.

A result of all the tested application was the investment of hundred of hours in data

analysis which yielded a wealth of experience in the operation of LIDAR processing software.

From that experience it can be concluded that there is no single software package that provides

all possible tools required for the application of TLS dataset for scientific purposes. For most of

the projects operations were performed in at least 3 software packages: Polyworks, Terrascan

and QT Modeler; in some cases Surfer was employed to create grids and Matlab was used to

code special customized codes for the extraction of the required final information.

8.2 Recommendations

There is a need to develop methods to improve the accuracy of the merging and geo-

referencing procedures. Extensive research should be performed by future students on those

fields as well as the establishment of a set of accuracy standards that will provide a benchmark

for measuring the quality of these operations.

It is also necessary to work in the field of sensor characterization, which includes the

development of test protocols to verify TLS specifications such as accuracy (under various

ranges, target reflectance and scan angles), scan speed and measurement repeatability. This

activity also includes the identification and characterization of systematic errors for individual

units. If a good test protocol is developed it will set the ground for an independent experiment to

test and compare different makes and models of TLS units which can provide material for a great

engineering paper.









The development of the Mobile Terrestrial Laser Scanning system must be accelerated,

more students must be involved in the project and they can participate more actively in the

integration of the different system subcomponents.

One of the lessons learned from this work is that the multidisciplinary background of the

students from both the GSE and ASPL groups is an invaluable academic and research asset.

However more actions have to be taken towards increasing the synergies between its members

with respect to the application of MTLS to diverse scientific disciplines.

Networking activities must be undertaken within the UF academic community to identify

potential partners that can use TLS systems in their research. Collaborative research projects

should be started; the results can yield publishable material.

Some consideration should be given to develop a program such as NCALM that will make

the M-TLS system and services available to scientific community. This program will not only

increase system usage, but also increase the operator experience, the capability testing and

provide more opportunities and material to publish.

Efforts should be made to participate in TLS related conferences such as the International

Society for Photogrammetry and Remote Sensing Commission III on Photogrammetric

Computer Vision and Image Analysis, where a great part of the published work refers to

processing and interpretation of laser range data. Also an annual meeting from an engineering

and industry perspective is organized by Spar Point Research LLC; this meeting focuses on

advanced dimensional control, work processes and 3D laser scanning technologies. The

participation in events of this sort will help UF researchers to keep on the leading edge of

scientific and engineering applications of Terrestrial Laser Scanning systems.









APPENDIX
COMPARISON OF TERRESTRIAL LASER SCANNERS

A.1 Optech ILRIS 3D

This is a dual scanning mirror, pulsed infrared laser system with the highest range

available on the market: from 3 m to beyond 1 km. It has an integrated megapixel digital camera

and large-format LCD viewfinder. 400 x 400 instantaneous field of view. An optional Pan and

Tilt base allows for a panoramic -20 to 900 x 3600 scanning coverage. Web site Reference:

http://www.optech.ca/i3dfeat-ilris.htm










Figure A-1. Optech ILRIS TOF TLS

A.2 Leica HDS3000

Leica High-Definition Surveying 3000, Nd:YAG frequency doubled pulsed laser, with a

single mirror scanner, with a dual-window and rotating base that allows for a 3600 x 2700 field

of view. Has an integrated Bore-sighted Single 240 x 240, 1024 x 1024 digital camera for

automatically calibrated photo overlays. Web site Reference: http://www.leica-

geosystems.com/hds/en/lgs_5574.htm










Figure A-2. Leica HDS 3000 TOF TLS









A.3 Leica HDS4500 25 & 53m

The Leica HDS4500 is an ultra-high speed 690 nm laser, phase-based, short range scanner.

Capable of scanning from 100,000 points/sec to 500,000 points/sec on a 3600 x 3100 field of

view. Two different models are capable of 25 and 53 m ambiguous ranges. Web site Reference:

http://www.leica-geosystems.com/hds/en/lgs_5572.htm












Figure A.-3. Leica HDS 4500 PD TLS

A.4 RIEGL LMS-Z420i

Designed as a High-Accuracy & Long-Range pulsed laser, single vertical line rotating

polygon scanner with a 800 range. The azimuth scan is accomplished by a rotating base full

3600. An optional digital camera can be mounted for photorealistic textured point clouds. Range

up to typ. 1000 m, precision up to 10 mm and scan rate up to 12 000 pts/s. Web site Reference:

http://www.riegl.com/terrestrial_scanners/lms-z420i_/420i_all.htm


Figure A-4. Riegl LMS-Z420i TOF TLS









A.5 RIEGL LMS-Z390

Designed as a High-Accuracy & High Resolution pulsed laser, single vertical line rotating

polygon scanner with a 800 range. The azimuth scan is accomplished by a rotating base full

3600. An optional digital camera can be mounted for photorealistic textured point clouds. Range

up to typ. 300 m, precision up to 2 mm and scan rate up to 12 000 pts/s. Web site Reference:

http://www.riegl.com/terrestrial_scanners/lms-z390_/390_all.htm












Figure A-5. Riegl LMS-Z390 TOF TLS

A.6 RIEGL LMS-Z210ii

Designed as a general purpose pulsed laser, single vertical line rotating polygon scanner

with a 800 range. The azimuth scan is accomplished by a rotating base full 3600. An optional

digital camera can be mounted for photorealistic textured point clouds. Range up to typ. 650 m,

precision up to 10 mm and scan rate up to 12 000 pts/s. Web site Reference:

http://www.riegl.com/terrestrial_scanners/lms-z21 ii/21 Oii_all .htm








Figure A-6. Riegl LMS-Z210ii TOF TLS


Figure A-6. Riegl LMS-Z210ii TOF TLS









A.7 Trimble GS101

A single scan line pulsed green laser system, with an integrated digital camera, on a

rotating base. It allows to scan in 600 vertical and 3600 horizontal, up to a range of 200 -350 m,

with a spot spacing of 32 iprad and a range resolution of 1.5 mm @ 50m and a scan rate of up to

5,000 pts/s. Web site Reference: http://www.trimble.com/gs200.shtml


Figure A-7. Trimble GS101 TOF TLS


A.8 Trimble GX 3D

A single scan line pulsed green laser system, with an integrated digital camera, on a

rotating base. It allows to scan in 600 vertical and 3600 horizontal, up to a range of 200 -350 m,

with a spot spacing of 32 ptrad and a range resolution of 1.5 mm @ 50m and a scan rate of up to

5,000 pts/s. Web site Reference: http://www.trimble.com/trimblegx.shtml


Figure A-8. Trimble GX 3D TOF TLS








A.9 Minolta VIVID 910
The VI-910 uses LASER triangulation. The entire area is captured in 2.5 seconds, and the

surface shape is converted to a lattice of over 300,000 vertices (connected points). A (24-bit)

color image is captured at the same time by the same triangulation CCD. The range is limited to

2.5 meters and the range resolution is in the order or 8 micrometers. Web site Reference:

http://www.minolta3d.com/products/main-en.asp





*Q ,
~lo




Figure A-9. Minolta VIVID 910 OT TLS
A.10 Zoller-Frohlish IMAGER 5006
Is an ultra-high speed visible laser, phase-based, short range scanner. Capable of scanning

from up to 500,000 points/s on a 3600 x 3100 field of view with an Ambiguity range up to 79

meters. Web site Reference: http://www.zf-laser.com/eimager5006.html










Figure A-10. Zoller-Frohlish IMAGER 5006 PD TLS











A.11 IQSun 880

Is an ultra-high speed visible 785 nm laser, phase-based, single line short range scanner.

Capable of scanning from up to 240,000 points/s on a 3600 x 3200 field of view with an

Ambiguity range up to 76 meters.


Figure A-11. IQSun 880 PD TLS

A.12 Comparison of Terrestrial LASER Scanner Specifications

The specifications of commercial TLS units was compiled and presented in Table A-1.

Table A-1. Comparison of terrestrial laser scanner specifications


Manufacturer
Model
Method of operation
Range [m]



Range resolution [mm]
Azimuth, elevation
resolution []
IFOV [Ver X Hor]
Aux FOV [Ver X Hor]
Laser type/color
Laser wavelength [nm]
Scan rate [points/s]
Beam divergence []
Texture
Weight [kg]
Dimensions (LxWxH) [cm]
Power Supply

Power consumption [W1


Optech
ILRIS 3D
Time of Flight
3-1500 @80%
3-800 @220%
3-350 (@4%


0.00115
40 x 40
-20 to 90 x 360
Infrared
1500
2,000
0.00974
Intensity & RGB
12
32 x 32 x 22

24 VDC

75


Leica
HDS3000
Time of Flight

300 @090%
134 (@18%


0.00022918


270 x 360
Visible Green
532
4,000
0.00687549
Intensity & RGB
17
26.5 x 37 x 51

36 V, AC or DC

80


Leica
HDS4500 25m
Phase Difference

0.1-25.2

3
20.0535228


310 x 360
Visible Red
690
500,000
0.02864789


18 x 30 x 35
24V DC
90 260V AC
50-70


Leica
HDS4500 53m
Phase Difference

0.1-53.5

5
20.0535228


310 x 360
Visible Red
690
500,000
0.02864789


19 x 30 x 35
24V DC
90 260V AC
50-71












Table A-1. Continued
Manufacturer RIEGL RIEGL RIEGL Trimble
Model LMS-Z420i LMS-Z390 LMS-Z210ii GS101
Method of operation Time of Flight Time of Flight Time of Flight Time of Flight
Range [m]
2-1000 @80% 1-300 @80% 4-650 @80%
2-350 0@10% 1-100 @10% 4-200 2@10%

Range resolution [mm] 10? 6? 15? 1.4-6.5
Azimuth, elevation
Azimut elevation 0.002, 0.0025 0.001 0.005 1.83346494
resolution [0]
IFOV [Ver X Hor] 80 x 0.014 80 x 0.014 80 x 0.15
Aux FOV [Ver X Hor] 80 x 360 80 x 360 80 x 360 60 x 360
Laser type/color Near Infrared Near Infrared Near Infrared Green
Laser wavelength [nm]
Scan rate [points/s] 8,000-12,000 8,000-11,000 8,000-12,000 5,000
Beam divergence [] 0.014 0.014 0.15 0.00343775
Texture Intensity & RGB Intensity & RGB Intensity & RGB Intensity & RGB
Weight [kg] 14.5 14.5 13 12.8
Dimensions (LxWxH) [cm] 21 x 21 x 46.3 21 x21 x 46.3 21 x 21 x 43.7 34 x 27 x 42
Power supply 90-240 VAC
12-18 VDC 12-18 VDC 12-18 VDC 9C
24 VDC
Power consumption [W] 78-98 78-98 78-98

Table A-1. Continued
Manufacturer Trimble Trimble Minolta Zoller-Frohlish
Model GS200 GX 3D VIVID 910 IMAGER 5006
Optical
Time of Flight Time of Flight Oticl Phase Difference
Method of operation Triangulation
Range [m]
2-350 350 @090% 0.6 -2.5 79

Range resolution [mm] 1.4-6.5 7 0.008 0.1
Azimuth, elevation
Azimuth, elevation 0.00171887 0.00343775 0.0018
resolution [0]
IFOV [Ver X Hor]
Aux FOV [Ver X Hor] 60 x 360 60 x 360 310 x 360
Laser type/color Green Green Visible Red Visible
Laser wavelength [nm] 532 690
Scan rate [points/s] 5,000 5,000 120,000-256,000 500,000
Beam divergence [] 0.00343775 0.00343775 0.01260507
Texture Intensity & RGB Intensity & RGB none
Weight [kg] 12.8 13 11 14
Dimensions (LxWxH) [cm] 35 x 27 x 42 32.3 x 34.3 x 40.4 21.3 x 27.1 x 41.3 28.6 x 19 x 37.2
Power supply 90-240 VAC 90-240 VAC 1 t 2 V
100 to 240 VAC 24V DC
24 VDC 24 VDC
Power consumption [W] 100 60 50









LIST OF REFERENCES


Baltsavias, E.(1999) Airborne laser scanning: existing systems and firms and other resources.
ISPRS Journal ofPhotogrammetry & Remote Sensing 54, 164-198.

Baltsavias, E.P. (1999) Airborne laser scanning: basic relations and formulas. ISPRS Journal of
Photogrammetry & Remote Sensing 54, 199-214.

Bellian, J.A.; Jennette, D.C.; Kerans, C.; Gibeaut, J.; Andrews, J.; Yssldyk, B.; Larue, D. (2002)
3-Dimensional Digital Outcrop Data Collection and Analysis Using Eye-safe Laser
(LIDAR) Technology. Published by AAPG/Datapages, Inc., 1444 South Boulder, Tulsa,
OK, 74119, USA Retrieved March, 2007 from
http://www.searchanddiscovery.com/documents/beg3 d/index.htm.

Bender, Currie, Dickey, Eckhardt, Faller, Kaula, Mulholland, Plotkin, Poultney, Silverberg,
Wilkinson, Williams, Alley. (1973). The Lunar Ranging Experiment. Science, New Series,
182, (4109), 229-238.

Bennett, W.R., 2000. Background of an Inversion: The First Gas Laser. IEEE Journal on
Selected Topics In Quantum Electronics, 6 (6), 869-875.

Beraldin, J-A; Blais, F; Rioux, M; Domey, J; Gonzo,L; De Nisi,F; Comper,F; Stoppa, D;
Gottardi, M; Simoni,A.(2003) Optimized Position Sensors for Flying-Spot Active
Triangulation Systems. Proceedings of the Fourth International Conference on 3-D Digital
Imaging and Modeling (3DIM'03), October 06 10, 2003, Banff, Canada. 2003, 28- 36.

Bureau of Beaches and Coastal Systems (BBCC). Florida Department of Environmental
Protection (2004) Monitoring Standards for Beach Erosion Control Projects. Tallahassee,
Florida. Retrieved March, 2007 from http://www.dep.state.fl.us/beaches/.

Burkard, R.K. (1983). Geodesy for the Layman (Fourth Revision). Defense Mapping Agency
Technical Report TR-80-003. St. Louis, Missouri.

Callens, M., Verhoest, N.E.C. and Davidson, M.W. J.. (2006) Parameterization of Tillage-
Induced Single-Scale Soil Roughness from 4-m Profiles. IEEE Transactions on Geoscience
and Remote Sensing. 44 (4), 878 888.

Carter, W. (1973). The Lunar Laser Ranging Pointing Problem. Unpublished doctoral
dissertation, University of Arizona, Tucson.

Carter, W. & Shrestha, R. (1997). Airborne laser swath mapping: Instant snapshots of our
changing beaches. In Proceedings of the Fourth International Conference Remote Sensing
for Marine and Coastal Environments, Environmental Research Institute of Michigan
(ERIM), Ann Arbor, Michigan. p. 298-307.

Carter, W., Shrestha, R., Slatton, K.C., Cossio, T., Shrestha, K. (2005) Geodetic Laser Scanning:
Refractive Optics Offer Wide Variety of Scan Patterns. Poster at the American
Geophysical Union, Fall Meeting 2005, San Francisco, California.









Cheves, M. (1999). Geodimeter-The First Name in EDM. Professional Surveyor, 19 (2).

Cheves, M. (2002).In the Air, On the Ground: Laser-based Technology from Optech.
Professional Surveyor, 22 (10).

Curless, B.; Levoy, M. (1995). Better Optical Triangulation through Spacetime Analysis.
Procedings of the Fifth International Conference on Computer Vision (ICCV 95), June 20-
23, 1995, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA., 987-
994.

Davidson, M.W.J., Toan, T.L., Mattia, F., Satalino, G., Manninen, T., Borgeaud, M. (2000). On
the Characterization of Agricultural Soil Roughness for Radar Remote Sensing Studies.
IEEE Transactions on Geoscience and Remote Sensing, 38 (2), 630-640.

Defense Mapping Agency (DMA) (1991) Error Theory as applied to Mapping Charting and
Geodesy. DMA Technical Report 8400.1, Fairax, Virginia, USA.

Drake, D.R. (2002) Applications of Laser Scanning and Imaging Systems. Unpublished Masters
Thesis, University of Florida, Gainesville.

Ewing, Clair & Mitchell, Michael (1970). Introduction to Geodesy. Elsevier North Holland
Publishing Company, Inc, New York, New York, USA.

Flood, M. (2001) Laser Altimetry: From Science to Commercial LIDAR Mapping. Journal of
The American Society for Photogrammetry and Remote Sensing, 67 (11), 1209-1211.

Frohlich, C., Mettenleiter, M.(2004). Terrestrial Laser Scanning New Perspectives In 3d
Surveying. Proceedings of the ISPRS working group 8/2 'Laser-Scanners for Forest and
Landscape Assessment', Freiburg, Germany, October 3-6, 2004, 7-13.

Guenther, G., Goodman, L., Enabnit, D., Swift, R., Thomas, R.. (1978). Laser Bathymetry for
Near Shore Charting Application (Preliminary Field Test Results). Oceans, 10, 390- 396.

Hickman, D., Hogg, J. (1969) Application Of An Airborne Pulsed Laser For Near Shore
Bathymetric Measurements Remote Sensing Of Environ, 1 (1), P 47-58.

Hopkinson,C.; Chasmer, L.; Young-Pow, C.; Treitz, P. (2004) Assessing forest metrics with a
ground-based scanning LIDAR. Canadian Journal of Forest Research. 34, 573-583.

lavarone, A (2002) Laser Scanner Fundamentals. Professional Surveyor Magazine, 24 (5).

Institute of Electrical and Electronics Engineers (IEEE). (N.D.). Theodore H. Maiman. Published
by the IEEE Virtual Museum, New Brunswick, New Jersey, USA. Retrieved March, 2007,
from http://www.ieee-virtual-museum.org/collection/people.php?id=1234591&lid=1.

Institute of Electrical and Electronics Engineers (IEEE). (N.D.). Charles H. Townes. Published
by the IEEE Virtual Museum, New Brunswick, New Jersey, USA.Retrieved March, 2007,
from http://www.ieee-virtual-museum.org/collection/people.php?id=1234588&lid=1









Institute of Electrical and Electronics Engineers (IEEE). (N.D.). Arthur L. Schawlow. Published
by the IEEE Virtual Museum, New Brunswick, New Jersey, USA.Retrieved March, 2007,
from http://www.ieee-virtual-museum.org/collection/people.php?id=1234585&lid=1

Javan,A., Bennett, W.R., Herrott, D.R. (1961) Population inversion and continuous optical maser
oscillation in a gas discharge containing a He-Ne mixture. Phys. Rev. Lett., 6, 106-110.

Jenkins, J.C.; Chojnacky, D.C.; Heath, L.S.; Birdsey, R.A. 2003. National-scale biomass
estimation for United States tree species. Forest Science. 49(1), 12-35.

Kern, J. (2001) Mapping Groud Zero. Professional Surveyor, November 2001, 21 (10)

Kim, H; Cervenka, P; Lankford, C. (1975) Development of an Airborne Laser Bathymeter,
NASA Techincal Note D-8079, NASA Wallops Flight Center, Wallops Island, Virginia,
USA.

Kraus, N.C. & Galgano, F.A. (2001) Beach Erosional Hot Spots: Types, Causes, and Solutions.
US Army Engineer Research and Development Center/Coastal Hydraulics Laboratory.
Coastal and Hydraulics Engineering Technical Note ERDC/CHL CHETN-II-44.
Vicksburg, Mississippi, USA.

Kaufman, C.C., Batts, E.L., Levine, N.S., Doyle, B.C. (2005) Beach Erosion Mapping Using
LIDAR and a 3D Laser Scanner: Folly Beach, South Carolina. Poster presentation at the
Geological Society of America (GSA) Annual Meeting 2005, Salt Lake City, Utah,
October 16-19, 2005, paper 59-14

Lefsky,M.; Cohen, W.; Parker, W.; & Harding, D. (2002) LIDAR Remote Sensing for
Ecosystem Studies. BioScience, 52 (1), 19-30.

Lefsky, M.A.; Cohen, W.B.; Acker, S. A.; Spies, T.A.; Parke3, G.G.; Harding, D. (1998),
LIDAR Remote Sensing of Forest Canopy Structure and Related Biophysical Parameters
at the H.J. Andrews Experimental Forest, Oregon, USA. Geoscience and Remote Sensing
Symposium Proceedings, 1998. IGARSS '98, Seattle, WA, USA. 6-10 Jul 1998, 3, 1252-
1254.

Lichti, D.D.; Gordon, S.J. & Stewart, M.P.. (2002). Ground-based laser scanners: operation,
systems and applications. Geomatica, 56, 22-33.

Maling, D. H. (1992). Coordinate Systems and Map Projections (2nd Edition). Pergamon,
Oxford, United Kindom.

Marino, R.; Davis, W.R. (2005) Jigsaw: A Foliage-Penetrating 3D Imaging Laser Radar System.
Lincoln Laboratory Journal, 15 (1), 23-36.

Martel, S.J. (2006) Effect of topographic curvature on near-surface stresses and application to
sheeting joints: Geophysical Research Letters, 33 (1) pg:L01308.









Mattia, F.; Davidson, Malcolm W.J.; Toan, Thuy Le; D'Haese, Christophe M. F.; Verhoest, Niko
E.C.; Gatti, Anna Maria; Borgeaud, Maurice. (2003) A Comparison Between Soil
Roughness Statistics Used in Surface Scattering Models Derived From Mechanical and
Laser Profilers. IEEE Transactions on Geoscience and Remote Sensing, 41 (7), 1659-1671.

McClung, F.J. and Hellwarth, R.W. (1962): "Giant optical pulsations from ruby". Journal of
Applied Physics, 33 (3), 828-829.

McGarry, Jan; Zagwodzki, Tom. (2005) A Brief History of Satellite Laser Ranging: 1964 -
present*. Published by the Crustal Dynamics Data Information System (CDDIS), NASA
Goddard Space Flight Center, Greenbelt, Maryland. Retrieved October, 2005, from
http://cddis.nasa.gov/slr2000/docs/gsfcslrjm0504.pdf

Mitchell, R (N.D.) AOL History WFF AOL. Published by the Hydrospheric and Biospheric
Sciences Laboratory, NASA GSFC/Wallops Flight Facility, Wallops Island, Virginia, USA
Retrieved March, 2007, from
http://aol.wff.nasa.gov/index.php?module=pagemaster&PAGEuser op=viewpage&PAG
E_id=3&MMNposition=3:3.

Naesset, E & Gobakken, T. (2005) Estimating forest growth using canopy metrics derived from
airborne laser scanner data. Remote Sensing of Environment 96, 453 465.

National Aeronautics and Space Administration (NASA). (N.D.) National Space Science Data
Center Apollo Laser Altimeter Experiment. Published by the NASA Goddard Space Flight
Center, Greenbelt, Maryland. Retrieved March, 2007, from
http://nssdc.gsfc.nasa.gov/database/MasterCatalog?sc=1971-063A&ex=5.

National Oceanic and Atmospheric Administration (NOAA), (N.D.). Distance Measurement
Tools, AGA Geodimeter NASM-2A. Published by the NOAA the Office of Public Affairs,
Washington, D.C., USA. Retrieved March, 2007, from
http://celebrating200years.noaa.gov/distancetools/0411 .html.

National Oceanic and Atmospheric Administration (NOAA), (N.D.). Distance Measurement
Tools, Tellurometer. Published by the NOAA the Office of Public Affairs, Washington,
D.C., USA.Retrieved March, 2007, from
http://celebrating200years.noaa.gov/distancetools/tellurometermral.html.

National Oceanic and Atmospheric Administration (NOAA), (N.D.). Distance Measurement
Tools, Ranger III and Rangemaster III. Published by the NOAA the Office of Public
Affairs, Washington, D.C., USA.Retrieved March, 2007, from
http://celebrating200years.noaa.gov/distancetools/ranger.html.

Oelze, Michael L.; Sabatier, James M.; Raspet, Richard (2003). Roughness Measurements of
Soil Surfaces by Acoustic Backscatter. Soil Science Society of America Journal, 67, 241-
250.









Parrish, C.E.; Tuell, G.H.; Carter,W.E. and Shrestha, R.L. (2005) Configuring an Airborne Laser
Scanner for Detecting Airport Obstructions. Photogrammetric Engineering & Remote
Sensing, 71 (1), 37-46.

Perron, Jay Taylor (2006). Formation of Evenly Spaced Ridges and Valleys. Unpublished
doctoral dissertation, University of California, Berkeley.

Schawlow, A. L. & Townes, C. H. (1958) Infrared and Optical Masers. Physical Review, 112,
1940-1949.

Shrestha, Ramesh (1983). Local Geodetical Networks Adjustments in Three Dimensions.
Unpublished doctoral dissertation, University of Wisconsin-Madison.

Shrestha,R.L., Carter,W.E., Sartori,M., Luzum, B.J., Slatton, K.C.(2005) Airborne Laser Swath
Mapping: Quantifying changes in sandy beaches over time scales of weeks to years. ISPRS
Journal of Photogrammetry & Remote Sensing, 59, 222-232.

Smith, James R. (1996). Introduction to Geodesy. John Wiley & Sons, Inc. New York, New
York, USA.

Smithsonian National Museum of American History. Virtual Surveying Instrument Collection.
(ND). EDM (Geodimeter Model 2A). Published by the Smithsonian National Museum of
American History. Washington. D.C., USA. Retrieved March, 2007, from
http://americanhistory.si.edu/collections/surveying/object.cfm?recordnumber=748815.

Smithsonian National Museum of American History. Virtual Surveying Instrument Collection.
(ND). EDM (Tellurometer M/RA 1). Published by the Smithsonian National Museum of
American History. Washington. D.C., USA. Retrieved March, 2007, from
http://americanhistory.si.edu/collections/surveying/object.cfm?recordnumber=759173

Smithsonian National Museum of American History. Virtual Surveying Instrument Collection.
(ND). EDM (Tellurometer M/RA 1). Published by the Smithsonian National Museum of
American History. Washington. D.C., USA. Retrieved March, 2007, from
http://americanhistory.si.edu/collections/surveying/maker.cfm?makerid=45

Smithsonian National Museum of American History. Virtual Surveying Instrument Collection.
(ND). EDM (Geodolite). Published by the Smithsonian National Museum of American
History. Washington. D.C., USA. Retrieved March, 2007, from
http://americanhistory.si.edu/collections/surveying/object.cfm?recordnumber=748473

Taylor, Nick (2000). LASER: The inventor, the Nobel laureate, and the thirty-year patent war.
Simon & Schuster. New York, New York, USA.

Thoma,D. P., Moran,M.S., Bryant,R., Rahman,M., Holifield-Collins,C.D., Sano,E.E., Slocum,
K., Skirvin,S..(2006) Comparison of four models to determine surface soil moisture from
C-band radar imagery in a sparsely vegetated semiarid landscape. Water Resources
Research, Vol. 42, W01418









Watt, P. J. & Donoghue, D. N. M. (2005) Measuring forest structure with terrestrial laser
scanning. International Journal of Remote Sensing, 26 (7), 1437-1446.

Wehr, A; Lohr U. (1999) Airborne Laser Scanning-An Introduction and Overview. ISPRS
Journal of Photogrammetry & Remote Sensing 54, 68-82.

Zribi,M., Ciarletti, V., Taconet, O., Boissard, P., Chapron, M. and Rabin, B. (2000)
Backscattering on soil structure described by plane facets. International Journal on
Remote Sensing, 21 (1), 137-153.









BIOGRAPHICAL SKETCH

Juan Carlos Fernandez Diaz was born August 18, 1976, in Tegucigalpa, Honduras, to

Venancio Fernandez and Ana Maria Diaz. His family is also composed by one sister: Maria

Esther and two brothers David and Jose Venancio. Since a very young age he developed strong

interest towards science and technology; especially to earth and space science, aviation,

telecommunications and electronics. He was fortunate to attend an American school (Elvel

School) from kindergarten to 11th grade, where the professors motivated his scientific curiosity.

He graduated from High School in 1993 from a program that fulfills the requirements from both

the American and Honduran academic curriculum. That same year he enrolled on the Electrical

Engineering program of the Universidad Nacional Autonoma de Honduras (UNAH). When he

did not find college challenging enough he decided to work full time while pursuing the B.S.

degree. His first position was as an instructor of the university's astronomical observatory where

he acquired knowledge and expertise related to the design, use and maintenance of astronomical

instrumentation as well as astronomical data processing and analysis. During this period he also

participated in a train ship at the European Space Agency Satellite Tracking Station in

Villafranca del Castillo, Spain and received a Summer Undergraduate Research Fellowship

(SURF) from the California Institute of Technology (CALTECH) to perform scientific research

at the Jet Propulsion Laboratory. Soon after this experience he accepted a new position with the

Honduras National Telecommunication Commission as a Spectrum Planning and Engineering

technician. He obtained the BS degree in electrical engineering in June 2001 having not only the

formal academic knowledge but also a great deal of experience in telecommunications, space

science and technology.

From 2002 to 2005 he continued his career in telecommunications holding positions in a

Wireless service provider were he performed functions as network planning engineer and quality









assurance chief. He also obtained a Master of business administration degree with a summa cum

laude distinction from the Universidad Catolica de Honduras in 2005. During 2004 he applied

for a Fulbright Scholarship to participate in a masters program in the fields of Satellite

Applications (Navigation, Communications and Remote Sensing). He was fortunate to receive

the scholarship and to be accepted to the University of Florida, Geosensing Systems Engineering

graduate program.

Juan hopes to continue with his multidisciplinary education, continue to explore his

interest in space science and technology and to contribute in some way to mankind progress,

especially to the benefit of his fellow Hondurefios.





PAGE 1

1 SCIENTIFIC APPLICATIONS OF THE MOBILE TERRESTRIAL LASER SCANNER (M-TLS) SYSTEM By JUAN CARLOS FERNANDEZ DIAZ A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Juan Carlos Fernandez Diaz

PAGE 3

3 To my parents and to my siblings. Who I am and what I have achieved is a result of your dedication, hard work and motivation. Also, to all the close friends who believed in me, encouraged me, and provided their support.

PAGE 4

4 ACKNOWLEDGMENTS First, DEO GRATIAS. Second, I want to ac knowledge the great support and guidance from Ramesh Shrestha, William Carter and K. C lint Slatton. I thank them for sharing with me their knowledge, experience, and spirit, and for their continued motivation and commitment to make the geosensing gradua tes leaders in the field. I gratefully acknowledge the pe rsons and institutions that make the J.William Fulbright Foreign Scholarship program a reality. Through th em I had the unique opportunity to achieve higher education in one of the top universi ties of the United States of America. To all my friends, classmates and colleague s from the Geosensing Systems Engineering program and the NSF supported National Center for Airborne Laser Mapping, who provided me with invaluable support to complete this work. Finally to the countless persons and institut ions who in some way or another have contributed to my success in the UF Geosensing Graduate Program.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES................................................................................................................ .......10 ABSTRACT....................................................................................................................... ............14 1 INTRODUCTION TO TERREST RIAL LASER SCANNING.............................................16 1.1 Geodesy and the Need for Measurements......................................................................16 1.2 History of LASER EDM and Scanners..........................................................................19 1.3 TLS Subsystems and Pr inciples of Operation................................................................29 1.3.1 LIDAR Ranging Principles..................................................................................29 1.3.1.1 Phase difference measurement (PD)..........................................................30 1.3.1.2 Time of flight (TOF)..................................................................................32 1.3.1.3 Optical triangulation (OT)..........................................................................33 1.3.2 Scanning Mechanisms..........................................................................................34 1.4 Technical Characteristic s and Specifications of TLS.....................................................35 1.4.1 LASER Type........................................................................................................35 1.4.2 LASER Class........................................................................................................36 1.4.3 LASER Beam Divergence....................................................................................36 1.4.4 Point Spacing........................................................................................................37 1.4.5 Range.................................................................................................................. ..37 1.4.6 Range Resolution..................................................................................................37 1.4.7 Precision and Accuracy........................................................................................37 2 UF MOBILE TERRESTRIAL LA SER SCANNING (M-TLS) SYSTEM...........................39 2.1 Evolution of the M-TLS Concept...................................................................................39 2.2 M-TLS Subsystems........................................................................................................4 1 2.2.1 LIDAR Unit.........................................................................................................41 2.2.2 Vehicle............................................................................................................... ..44 2.2.3 Lift.................................................................................................................. .....45 2.2.4 Power Subsystem.................................................................................................45 2.2.5 Pan Tilt Base........................................................................................................4 6 2.2.6 Video Camera......................................................................................................47 2.2.7 On Board PC........................................................................................................47 2.2.8 GPS................................................................................................................... ...47 2.2.9 Tiltmeter............................................................................................................. .49 2.2.10 INS.................................................................................................................... ...49

PAGE 6

6 3 WORKFLOW OF M-TLS OPERATIONS............................................................................50 3.1 Data Collection.......................................................................................................... .....50 3.2 Data Parsing............................................................................................................. .......51 3.3 Data Manipulation and Information Extraction.............................................................53 3.3.1 Visualization.........................................................................................................5 3 3.3.2 Single Point Selection...........................................................................................54 3.3.3 Measurements.......................................................................................................54 3.3.4 Primitive Fitting....................................................................................................54 3.3.5 Generating Cross Sections....................................................................................55 3.3.6 Transformations....................................................................................................56 3.3.6.1 Rotations and translations..........................................................................56 3.3.6.2 Cropping.....................................................................................................56 3.3.6.3 Merging......................................................................................................56 3.3.6.4 Geo-referencing..........................................................................................57 3.3.7 Segmentation, Classification and Filtering...........................................................58 3.3.7.1 Segmentation..............................................................................................58 3.3.7.2 Classification..............................................................................................58 3.3.7.3 Filtering......................................................................................................58 3.3.8 Gridding............................................................................................................... .58 3.3.9 Advanced Mathematical Operations....................................................................59 4 TESTED APPLICATIONS OF M-TLS.................................................................................61 4.1 Common Applications of TLS........................................................................................61 4.2 Paleontology............................................................................................................. ......62 4.3 Structural Geology....................................................................................................... ...65 4.4 Wildlife Management Conservation...............................................................................69 4.5 Coastal Morphology....................................................................................................... 69 4.6 Soil Science............................................................................................................. .......71 4.7 Forestry................................................................................................................. ..........72 5 ST. AUGUSTINE BEACH EROSION HOT SPOT MAPPING...........................................74 5.1 Motivation............................................................................................................... ........74 5.2 Use of LIDAR Technology............................................................................................75 5.3 Data Collection.......................................................................................................... .....75 5.4 Data Processing.......................................................................................................... ....76 5.5 Results.................................................................................................................. ...........79 5.5.1 Elevation Changes................................................................................................79 5.5.2 Volume Changes..................................................................................................80 5.5.3 Beach Line and Crest of Berm Extraction From the Grids..................................80 5.5.4 Across Beach Profile Extraction..........................................................................81 5.6 Comparison Between Trad itional Methods and M-TLS................................................83

PAGE 7

7 6 SOIL ROUGHNESS METRICS DETERMINATION..........................................................84 6.1 Motivation............................................................................................................... ........84 6.2 Use of LIDAR Technology............................................................................................85 6.3 Data Collection.......................................................................................................... .....86 6.4 Data Processing.......................................................................................................... ....87 6.5 Results.................................................................................................................. ...........89 6.5.1 Simulated Profiling Results..................................................................................89 6.5.2 Extension of the 2D Formulas for a 3D Surface..................................................92 6.5.3 Comparisons of Roughness Metric s From Profiles vs. Full Surface....................95 6.5.4 Distribution Functions of the 3D Correlations Lengths......................................95 6.6 Comparison with the Tr aditional Meshboard Method....................................................96 6.7 Conclusions.............................................................................................................. .......98 7 FORESTRY METRICS APPLICATIONS............................................................................99 7.1 Motivation............................................................................................................... ........99 7.2 Use of LIDAR Technology..........................................................................................100 7.3 Previous Works........................................................................................................... ..101 7.4 Data Collection.......................................................................................................... ...102 7.5 Data Processing.......................................................................................................... ..104 7.6 Results.................................................................................................................. .........108 7.6.1 Stem Density......................................................................................................108 7.6.2 Stem Location.....................................................................................................110 7.6.3 Stem Diameter at Breast Height DBH................................................................110 7.6.4 Tree Height.........................................................................................................111 7.6.5 Stem Volume......................................................................................................112 7.6.6 Tree Biomass Estimation....................................................................................113 7.7 Comparison with Traditional Methods.........................................................................113 7.8 Conclusions.............................................................................................................. .....114 8 SUMMARY...................................................................................................................... ....116 8.1 Conclusions.............................................................................................................. .....116 8.2 Recommendations.........................................................................................................1 17 APPENDIX COMPARISON OF TE RRESTRIAL LASER SCANNERS.................................119 A.1 Optech ILRIS 3D.........................................................................................................1 19 A.2 Leica HDS3000............................................................................................................ 119 A.3 Leica HDS4500 25 & 53m..........................................................................................120 A.4 RIEGL LMS-Z420i......................................................................................................120 A.5 RIEGL LMS-Z390.......................................................................................................121 A.6 RIEGL LMS-Z210ii....................................................................................................121 A.7 Trimble GS101............................................................................................................ 122 A.8 Trimble GX 3D............................................................................................................122 A.9 Minolta VIVID 910.....................................................................................................123

PAGE 8

8 A.10 Zoller-Frohlish IMAGER 5006.................................................................................123 A.11 IQSun 880............................................................................................................... ...124 A.12 Comparison of Terrestrial LASER Scanner Specifications.......................................124 LIST OF REFERENCES............................................................................................................. 126 BIOGRAPHICAL SKETCH.......................................................................................................132

PAGE 9

9 LIST OF TABLES Table page 1-1. Phase difference ranging princi ples expressions and equations........................................31 1-2. Time of flight ranging prin ciples expressions and equations............................................33 1-3. Optical triangulation ranging pr inciples expressions and equations..................................34 1-4. LASER classification..................................................................................................... ....36 5-1. Control points used for the geo-re ferencing of the March 23, 2006 dataset......................77 5-2. Point clouds coregistration RMS values............................................................................77 5-3. Summary of volume change computations........................................................................80 6-1. Soil roughness collected datasets.......................................................................................8 6 6-2. Definition of soil roughness parameters............................................................................89 6-3. Soil roughness parameters results from random profiles for the Citra 02 grid..................91 6-4. Soil roughness parameters results from random profiles for the Hastings grid.................91 6-5. Soil roughness parameters from 3D surface models..........................................................95 6-6. Comparison of soil roughness parameters for Citra 02 from 3D surface models and random generated profiles..................................................................................................95 6-7. Comparison of soil roughness parameters for Hastings 02 from 3D surface models and random generated profiles...........................................................................................95 6-8. Comparison of soil roughness metrics obtai ned from the traditional and alternative method......................................................................................................................... .......97 7-1. M-TLS data set geo-re ferencing control network............................................................104 7-2. Geo-referencing residuals analysis..................................................................................105 7-3. Diameter and heights measurem ents for stem volume estimation...................................113 A-1. Comparison of terrestrial laser scanner specifications.....................................................124

PAGE 10

10 LIST OF FIGURES Figure page 1-1. Triangulation network.................................................................................................... ....17 1-2. Early EDM equipment...................................................................................................... .20 1-3. K+E RangeMaster III EDM unit........................................................................................24 1-4. Spectra-Physics Geodolite................................................................................................ .25 1-5. NASA AOL system.......................................................................................................... .27 1-6. Phase difference ranging principle.....................................................................................31 1-7. Time of flight ranging principle......................................................................................... 32 1-8. Optical triangulat ion ranging principle..............................................................................33 1-9. Scanning axes on a panoramic view scanner.....................................................................34 2-1. Survey vehicle with a similar concept to MOBLASS.......................................................41 2-2. OPTECH ILRIS block diagram.........................................................................................42 2-3. Mobile Terrestrial Laser Scanning (M-TLS) system: truck, lift and ILRIS......................44 2-4. M-TLS deployed in Georgia, showing th e power generator and field computer on the truck bed...................................................................................................................... .......45 2-5. M-TLS pan tilt bases..................................................................................................... .....46 2-6. The ILRIS unit with on axis video camera and GPS antenna............................................48 2-7. Array of geodetic quality GPS base stations used for geo-referencing.............................48 2-8. Installation of the tiltmeter unit on the instrument frame..................................................49 3-1. Screen capture of the ILRIS unit cont roller software duri ng scanning operations............51 3-2. Different configurations of the MTLS system during seve ral data collection projects....................................................................................................................... ........51 3-3. Screen capture of the Parser softwa re during the setup of parsing settings.......................52 3-4. Visualization of LIDAR data.............................................................................................5 4 3-5. Primitive fitting process illustrated.................................................................................... 55

PAGE 11

11 3-6. Cross section generation from the point cloud..................................................................55 3-7. Point cloud merging example............................................................................................57 3-8. Gridding operations, fr om point cloud to grid...................................................................59 3-9. Examples of advanced mathematical operations in the processing of Airborne LIDAR data..................................................................................................................... ...60 4-1. Traditional and alternative methods for measuring and recordi ng spatial information in archeological and paleontological sites.........................................................................63 4-2. Rendering of the dig site point cloud m acro model showing laser return intensity..........64 4-3. RGB textured renderings of high resolution point clouds.................................................64 4-4. 3D surface grids used to compute th e volume of dirt extracted in one day.......................65 4-5. Traditional and alternative ways to perform geologi cal field mapping.............................66 4-6. Geo-referenced point cloud re ndering of the Tuolumne quarry........................................68 4-7. Shaded relief image from a gridde d model of the south wall of the quarry......................68 4-8. Methods of measuring alligators and crocodiles...............................................................69 4-9. Methods for generating beach profiles..............................................................................70 4-10. Methods for derivi ng soil roughness metrics.....................................................................71 4-11. Methods for estimating forestry metrics............................................................................72 5-1. Beach erosion hot spot study site location.........................................................................76 5-2. Rendering of the March 23rd, 2006 dataset........................................................................77 5-3. Features used to check the co registration of the point clouds..........................................78 5-4. Image maps from th e 10 cm elevation grids......................................................................78 5-5. Image maps from the elevation change grids....................................................................79 5-6. Beach line and crest of berm position plot s for each of the data collection dates.............81 5-7. Beach profiles extracted from the grids showing the recession of the berm.....................82 5-8. Comparison of profile resolution gene rated from traditional methods and M-TLS..........83 6-1. Dataset preprocessing steps.............................................................................................. .87

PAGE 12

12 6-2. Renderings of the 1 cm elevation grids..............................................................................88 6-4. Roughness parameter plots for the Citr a 02 dataset parallel to the X axis........................90 6-5. Roughness parameter plots for the Citr a 02 dataset parallel to the Y axis........................90 6-6. Roughness parameter plots for the Hasti ngs dataset parallel to the X axis.......................90 6-7. Roughness parameter plots for the Hasti ngs dataset parallel to the Y axis.......................91 6-8. Citra normalized he ight autocorrelation............................................................................92 6-9. Hastings height autocorrelation.........................................................................................9 3 6-10. Correlation lengths extrac tion for the Citra 02 dataset......................................................93 6-11. Correlation lengths extracti on for the Hastings dataset.....................................................94 6-12. Comparison of experimental correlation length distributions with respect to the assumed normal distribution..............................................................................................96 6-13. Meshboard used to digitize the soil surface transect.........................................................96 6-14. Plots of meshboard derived data........................................................................................9 7 7-1. Aerial photographs of the test site...................................................................................103 7-2. Shaded relief digital elevation model re ndered from the airborne laser scanner data of the test site............................................................................................................... ....104 7-3. Rendering of the fused point cloud, color coded by elevation........................................106 7-4. Rendering of fused point cloud cross s ection in the along the flightline direction..........106 7-5. Rendering of fused point cloud cross s ection in the cross flightline direction................106 7-6. Rendering of the fused point cloud, grey scale from the laser return intensity...............107 7-7. Rendering of the fused point cloud, color c oded by elevation + lase r return intensity...107 7-8. Rendering of the top view of fused poi nt cloud, color coded by elevation + laser return intensity............................................................................................................... ..108 7-9. Rendering of the “Forest Cube”.......................................................................................109 7-10. Rendering of point cloud used for stem counts...............................................................109 7-11. Fitting of a circle at breast hei ght for determining DBH and stem location....................110

PAGE 13

13 7-12. Single tree height determination......................................................................................11 1 7-13. Diameters at different he ights for volume computations.................................................112 7-14. Individual tree metric measurement.................................................................................114 A-1. Optech ILRIS TOF TLS..................................................................................................119 A-2. Leica HDS 3000 TOF TLS..............................................................................................119 A.-3. Leica HDS 4500 PD TLS................................................................................................120 A-4. Riegl LMS-Z420i TOF TLS............................................................................................120 A-5. Riegl LMS-Z390 TOF TLS.............................................................................................121 A-6. Riegl LMS-Z210ii TOF TLS...........................................................................................121 A-7. Trimble GS101 TOF TLS................................................................................................122 A-8. Trimble GX 3D TOF TLS...............................................................................................122 A-9. Minolta VIVID 910 OT TLS...........................................................................................123 A-10. Zoller-Frohlish IMAGER 5006 PD TLS.........................................................................123 A-11. IQSun 880 PD TLS........................................................................................................ ..124

PAGE 14

14 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science SCIENTIFIC APPLICATIONS OF THE MOBILE TERRESTRIAL LASER SCANNER (M-TLS) SYSTEM By Juan Carlos Fernandez Diaz August 2007 Chair: Ramesh Shrestha Cochair: K. Clint Slatton Major: Civil Engineering Terrestrial Laser Scanners (TLS) are ma pping instruments composed of a Laser rangefinder and an optical-mechanical system used to steer the laser be am across the surface of interest. The University of Florida Geosen sing Engineering and Mapping (GEM) Research Center is working towards developing a Mobile Terrestrial Laser Scanner (M-TLS) system. The core of the M-TLS is a commercia l 2-axis ground based laser sca nner which is integrated to a mobile telescoping, rotating, and tilting platform which provides up to 6 degrees of freedom for performing scanning operations. A scanner built-in 6 megapixel digital camera and a digital video camera provide the M-TLS moving and still imaging capability. At the final stage of development the M-TLS positioning and navigation system will include a differential GPS array, tilt sensors and an inertial measuring unit which will allow data collection and georeferencing in both static and dynamic modes. The M-TLS laser scanner is capable of genera ting 3D spatial and multispectral data sets. A typical dataset is composed of a cloud of milli ons of points for which 3D coordinates, laser intensity and/or RGB information are available for each one. Data can be collected in a range from 3m to 1500m for a target with an 80% refl ectivity or 3m to 350m to targets with a 4%

PAGE 15

15 reflectivity. The laser operates at a wavelength of 1535 nm, with a pulse width less than 10 ns and energy of less than 10 joules. The sampling separation can be adjusted down to 0.00115, and the scanning speed is 2,000 points per second. The M-TLS is a unique tool that enables GEM researchers to acquire high density point clouds from an advantageous terrestrial geom etry, being a very valuable complement for Airborne Laser Scanner data sets The applications of the M-TLS data sets are numerous in both the fields of science and engi neering. Tested applications by the GEM center include urban mapping, as-built surveying, building damage assessm ent, bridge load analysis, forestry metrics extraction, beach erosion mapping, paleontology a nd archeology dig mapping, structural geology mapping, forest fire fuel estima tion, soil spatial characterization and vehicle 3D modeling. This thesis centers around the novel applications of the M-TLS to specific scientific problems: thoroughly analyzing the applications to beach erosion hot spot mapping, soil roughness metrics extraction, and forestry metrics extraction.

PAGE 16

16 CHAPTER 1 INTRODUCTION TO TERRESTRIAL LASER SCANNING 1.1 Geodesy and the Need for Measurements Since early history mankind has had questi ons about nature, many of which can be answered if the proper measuremen ts are made. The shape and size of the Earth was one of the first questions that humans asked. The earliest answers to those que stions were based on folklore, common sense or primitive scientific method. These earliest conceptions had to do with supernatural creatures or deities and varied from culture to culture. The Greek philosophers did a lot of thinking about this issue. Homer, the epic Greek writer, popularized the conception that the Earth was a flat disk. It was around the time of Pythagoras, the Greek Mathematician who lived between 580 and 500 BC, that the idea that the Ea rth must be a sphere, the most perfect three dimensional geometric figure, took form. Fo llowing this Pythagorian principle other philosophers such as Aristotle and Archimedes tr ied to estimate the size of the Earth (Burkard, 1983; Smith, 1996). The first scientific approach to measure the si ze of the Earth is credited to Eratosthenes (Burkard, 1983; Ewing & Mitchell, 1970; Smit h, 1996), a Greek philosopher who also was a librarian in Alexandria. By st udying the difference in the length s of shadows cast by the Sun on the same day in the cities of Alexandria and Sy ene, and assuming that the Earth was a sphere, he was able to determine the angular separation be tween these two cities. Then by measuring the linear distance along the surface betw een the two cities he was able to estimate the size of the spherical Earth. The thoughts and works of thes e Greek philosophers led to the birth of the discipline that is called geodesy. Modern geodesy is defined as a branch of applied science that deals with the determination of the size, shape and orientation of the Earth and its gravitational field, and the variations of these parame ters with time (Ewing & Mitchell, 1970).

PAGE 17

17 A central focus of geodesy is determining the locations of points on the Earth, i.e., determining the coordinates of poi nts relative to a well defined or igin and reference frame. The evolution of surveying techniques and instrument ation led geodesists to treat positioning as two separate but related problems (Shrestha, 1983). For determining th e horizontal position components, astronomical observations, triangulati on, trilateration and traversing were some of the techniques employed. For determining the ve rtical position geodetic leveling, trigonometric leveling, barometric leveli ng or echo sounding techniques are employed (Smith, 1996). As described in this intr oduction, the foundations of geodesy are math and physics, however, these rely on real measurements to model the world. The most basic geodetic measurements are the ones pertaining to time, dist ances and angles. If clos e attention is given to the history of geodesy it becomes clear that gr eat leaps in knowledge came when new technology and its attendant instrumentation permitted leaps in the accuracy and precision of the measurements. Measuring the distance between widely separated points on Earth has always presented a challenge. One of the earliest succes sful methods of connecting widely separated points is a method known as triangulation. In a tr iangulation network, the a ngles of a chain of triangles are observed, along with one relatively short “baselin e” length. From this single distance observation, and using tr igonometry, all the sides of the triangles are computed. Figure 1-1. Triangulation network. The baselines A to B, as well as a ll the internal an gles need to be measured to determine the A-C distance.

PAGE 18

18 Triangulation is a time consuming method and errors in the angular and distance measurements propagate all along the networ k, reducing the precision and accuracy of the desired A to B distance. A breakthrough in distance measurements came in 1941, when a Swedish geodesist named Eric Bergstrand concei ved a new technique to measure the time it took a beam of light to travel a known distance to determ ine the speed of light. He then realized that if the speed of light were accura tely known, he could invert th at technique and measure the distance between two points (C arter, 1973). This was the da wn of Electronic Distance Measurement or EDM. Because of the curvature of the Earth, it still remained necessary to determine the distance between widely separate d points with a series of shorter distance measurements, in a method known as trilateration. The ease of use, accuracy and productivity of EDM instruments allowed trilateration to quickly displace triangulation. Since 1941 EDM has evolved from a concep t to a proven technology. Today there are many different forms of EDM instruments; some use radio frequencies, while others use light waves. EDM is used to measure distances small and astronomical, from micro structures to the distance between the Earth and the Moon, and even to neighboring planets. EDM has fulfilled the need for accurate distance measurements and has provided scientists and engineers with the data they need to build a model of our world. During the past decades EDM instruments have been developed that include optomechanical scanners which steer the measuring b eam over a selected pattern to collect a set of surface coordinates that can be us ed to create a mathematical representation of any surface in three-dimensional space. This thesis will present the experiences and results of almost two years of experimenting with a terrestrial LASER scan ner to fulfill the need for measurements in several scientific fields su ch as forestry, soil science, geology and beach morphology.

PAGE 19

19 1.2 History of LASER EDM and Scanners In tracing the origins of LASE R EDMs and scanners one can di g as deep as to the early astronomical observations aimed to estimate th e velocity of light, performed by Romer and Huygen, or the ground-based experiments conducte d by Hippolyte Fizeau, Leon Foucault, Simon Newcomb and Albert A. Michelson. However, most historians will set the origins of the EDM technique and instruments around the 1940s, highly influenced by the wartime efforts to develop the RADAR (Radio Detection and Ranging). The origin of the first EDM instrument be gan in 1938 when the physicist and geodesist Erik Bergstrand, of the Swedish Geographical Survey Office, began to investigate the possibilities of using a Kerr cell as an electro-optical sh utter to modulate a b eam of light in an attempt to measure of the speed of light. Bergst rand’s first operational instrument was reported to work in 1941. It used a Kerr Ce ll controlled by a crystal oscill ator to modulate light from an ordinary incandescent light bulb. The light beam was collimated and projected by a parabolic mirror to a reflective corner cube array, the returning waveform was detected by a photomultiplier tube (PMT), and the round trip travel time was determined from the difference in phase of the transmitted and refl ected modulated light beam (Carter, 1973; Smithsonian, N.D.). In 1947 Bergstrand took his instrument to a 6 km baseline in Orland and obtained a measurement of the speed of light of 299,793.1 0.2 km per second. A year later in August 1948, Bergstand read a paper at the meeting of the In ternational Association of Geodesy(IAG) held in Oslo, Norway. In that paper he explained that one could reverse the process, measure the light time of flight and use the value of the speed of light to compute th e distances between two points. Soon after the meeti ng, Bergstand licensed the concep t to the Swedish AGA (Svenska Aktiebolaget Gasacumulator) company to develop a commercial instrument.

PAGE 20

20 AGA produced the first EDM instrument in the early 1950s, and marketed it as the “Geodimeter”, short for “geodetic distance meter” The instrument used a Kerr cell to modulate the light, but the incandescent light bulb used by Bergstand was replaced with a mercury vapor light. The development of the Geodimeter by AGA continued through the 1950s and 1960s. The last model to use mercury lamps was the Geodimeter Model 6 introduced in 1964. Figure 1-2 shows an early production Geodime ter and its required corner cube reflector array. (NOAA, N.D.; Smithsonian, N.D.) Figure 1-2. Early EDM equipment. A) AGA Geodimeter B) Corner cube reflectors. (Source: http://pubs.usgs.gov/gip /monitor/techniques.html Last accessed March 16th, 2007) Around the same time that AGA was producing the first Geodimeter, Harry A. Baumann of the South African Trigonometrical Su rvey and Trevor Lloyd Wadley of the Telecommunications Research Laboratory of the South African Counc il for Scientific and Industrial Research (CSIR) were developing the Tellurometer. The Tellurometer was the first commercial EDM instrument to use microwave b eams to measure long distances with geodetic accuracy. The first model appeared in 1954, market ed as the Micro-Distancer M/RA 1. It was composed of two units, designated as master a nd remote, each set on the extreme points of the distance to be measured. Its range was between 30 and 50 km. The system used a continuous 3 GHz carrier frequency modulat ed by 10 megahertz and three ot her nearby frequencies. The

PAGE 21

21 remote station retransmitted the incoming wave in a similar wave of more complex modulation, and the resulting phase shifts of the carri er and the modulating signals was used to unambiguously determine the distance trav eled(NOAA, N.D.; Smithsonian, N.D.). The Tellurometer had the disadvantage that propaga tion of the microwave energy caused multipath reflections that degraded the system precision and accuracy. Between 1938 to 1960 EDM evolved from a concep t to widely used operational technique. However, its greatest leap in ra nge and accuracy was yet to be realized. In 1954, Charles Townes at the University of Columbia invented the MASER (Microwave Amplification by Stimulated Emission of Radiation). (IEEE, N.D.) A maser is a cavity filled with gas (the first used ammonia gas) that when “pumped” with microwave ra diation generates more microwave radiation. 1957 will always be remembered as the year that the Soviet Union launched the first artificial satellite – Sputnik I. But in Nove mber of 1957 Gordon Gould, a graduate student at Columbia University working with Townes, coined the acronym LASER, for Light Amplification by Stimulated Emission of Radiat ion, and described the pr incipal components of the LASER. However, Gould did not publish his wo rk, focusing his efforts in finding a position and the resources to try to build the LASER (IEEE, N.D.; Taylor, 2000). In March 1958, Arthur Schawlow, apparently inde pendently, also realized that the secret to the LASER involved an optical cavity, along th e lines of a Fabry-Pe rot interferometer. Schawlow shared his idea with Charles Townes, his brother-in-law, and together they wrote a paper entitled “Infrared and Optical Masers” publis hed in Physical Review Volume 112, number 6. (Schawlow & Townes, 1958) The paper by Schawlow and Townes encourag ed widespread thinking about how a Maser at optical wavelengths or LA SER might be built. In 1960 Theodore Maiman and his colleagues

PAGE 22

22 at Hughes Aircraft Company, succeeded in building the first solid state pulsed LASER, using a Ruby Rod. The LASER light was red with a wave length of 0.6943 micrometer s. That same year, Ali Javan and his colleagues from Bell Laboratories succeeded in building the first gas LASER. The HeNe LASER produced a continuous beam at five different wavelengths, achieving the highest power of 15 miliwatts at 1.153 microm eters (Javan et al., 1961 & Bennett, 2000). The gas LASER was well suited for use in terr estrial EDM instruments. The light produced by a LASER is highly mono-chromatic and coherent (the photons of a LASER beam have a single wavelength, phase and move in the same di rection). These attributes allow a LASER beam to have a small divergence, which means that the energy does not spread in the typical large spherical pattern of other light sources. Replacing the mercury vapor light with a HeNe LASER dramatically increased the opera ting range of the Geodimeter, and enabled the development of other smaller EDM instruments th at quickly took over many aspect s of surveying. The solid state pulsed ruby LASERs were not very energy efficient, and even th e best “Q-switched” devices produced light pulses meters in length, making them poorly suited for short range geodetic surveying, but they could produce large quantit ies of energy, however, they would soon find a role in space geodesy (Carter, 1973). For geodesists the main limitation of the improved LASER based EDM instruments was not the technical capabilities of the instruments themselves, but rather that the curvature and topography of the Earth limited the li ne of sight distances. As soon as the first artificial satellite was put in orbit in 1957, geodesist start to realize that a spacebor ne “target” could greatly extend the measured baseline distances. At the same time sc ientists interested in the fields of gravity and relativity (which also are of great importanc e to geodesy) started working on a concept to employ a high density and high altit ude artificial satellite to measure slow changes in the

PAGE 23

23 universal gravitation constant (G) by accurate tracking the satellite path using retroreflectors and pulsed search lights (B ender et al; 1973). In 1964 the first geodetic satellite (Beacon Explorer 22-B) wa s put in orbit, it had an array of corner cube reflectors that were illumi nated using pulsed ruby LASER beams, the first ranging measurements obtained on Oct 31, 1964 (C arter, 1973; McGarry & Zagwodzki, 2005). However, even before the satellite was launched scientist realized that low orbiting satellites imposed several challenges such as very shor t visibility times and Earth’s gravitational perturbations that would limit the qualit y of the relativist experiments. As early as 1962 J.E. Faller had proposed the id ea of placing a retroref lector on the surface of the Moon, which could be used to bounce back a LASER beam shot from the Earth. Between 1962 and 1964 experiments that included the de tection of LASER beams bounced from the moon’s surface were performed, and in 1965 th e Lunar Ranging Experiment (LURE) multiinstitutional team was formed. From 1965 to 1969 the LURE team focused their efforts to develop the largest and most sophi sticated EDM system to date. Their first great milestone was reached on July 21st, 1969, when Neil Armstrong aligned and leveled the first corner cube reflector array on the surface of the moon. Shortly after the installation of the array on the surface of the Moon, scientists on the Earth used the Lick Observatory’s 3.05 meter telescope and a Q switched pulsed ruby LASER to aim a 2 arc minute divergence beam to the array. The first successful return signal from the array was obtained on August 1, 1969 at Lick Observatory, shortly after, on August 20th the McDonald Observatory reported success obtaining returns. Succe ssful results were also reported that same year by the Air Force Cambridge Research La boratories (AFCRL) Lunar Ranging Observatory in Arizona (Bender et al., 1973; Carter, 1973).

PAGE 24

24 Down on Earth, AGA continued its development of the Geodimeter, introducing its Model 8 in 1967, which was its first instrument to us e a helium-neon LASER. The LASER allowed the extension of the measuring range of the lamp un its of 20 to 30 km to a range of 60 km in both day and night conditions. (Smithsonian, N.D. & Cheves, M; 1999) In 1965 LASER Systems & Electronics, Inc. wa s established by a team of physicists and engineers who had worked at the Engineering De velopment Center at Arnold Air Force Base in Tullahoma, Tennessee. In 1970, LASER Systems & Electronics unveiled thei r first electronic distance measuring instrument: the Ranger. This was the first competition AGA faced; it used a red visible LASER and was capable of ranging dist ances from 1 meter to 6 km with an accuracy of 5 mm +2 ppm. The EDM side of LASER Systems & Electronics was sold to Keuffel & Esser in 1971, which continued to manufactur e the Ranger, RangeMaster and AutoRanger (1977) series. (NOAA, N.D.; Smithsonian, N.D.) Figure 1-3. K+E RangeMaster III EDM unit. (Source: http://celebrating200years.noaa .gov/distance_tools/ranger.html last accessed March 20th, 2007) During the 1970’s several survey ing and electronic instrument companies developed EDM equipment. Among those companies were Cubic Co rporation, Hewlett-Packard, Wild and Zeiss. One of the trends of this period was to combin e an angular measuring device with an EDM into what was, and still is called a total station. These instruments continued to evolve into the simpler, more compact, accurate and cheaper unit s that can be found today. Most of these units

PAGE 25

25 require the use of corner cube reflectors to get a strong return signal that enable a computation of the distance. Some EDM devices followed a different evolutionary path, and they are able to use weak return signals that are bounced back from na tural targets such as the surface of the Earth or man made structures to compute the range from the instrument to the surface. Between 1964 and 1966 Spectra-Physics, a comp any based in Mountain View, California developed a series of precise LASER-based EDMs. Its first prot otype model, the Mark I, was designed as an airborne profile recorder (1964). The Mark II m odel was mounted in a Douglas A-26 and used to record a height profile across a stadium from a flying altitude of 300 meters. The Mark III model introduced in 1966 was mark eted as a “Geodolite”, its development was funded by the United States Army Engineer Topo graphic Laboratories. (Smithsonian, N.D.). This concept was further used not only as a profiler but also as a LASER altimeter and for bathymetric measurements. Figure 1-4. Spectra-Physics Geodolite. (Source: http://historywired.si.edu/object.cfm?ID=22 last accessed March 20th, 2007) In 1969 Hickman & Hogg were the first to report results on the feasibi lity of using a pulsed blue-green (frequency doubled NdYAG: Neodyni um Yttrium Aluminum Garnet) LASER from an airborne platform for near-shore beach re connaissance surveys over the shores of Lake Ontario. The concept was employed on the LASER altimeter experiment which was flown on the

PAGE 26

26 Apollo 15, 16 & 17 (1971, 1972 & 1972) to the Moon. The LASER altimeter experiment obtained data on the altitude of the Command Service Module (CSM) above the lunar surface. The altimeter was used to support mapping and panoramic camera photography. It operated in two modes. In the independent mode it perfor med ranging measurements every 20 sec. In the slave mode when the metric camera was operated, it automatically emitted a LASER pulse to correspond to a midframe range for each frame (NASA, N.D.). Also between 1971 and 1972 U.S. Nava l Oceanographic Office (NAVOCEANO) performed flight tests of a prot otype airborne LASER bathymetry profiler known as Pulsed Light Airborne Depth Sounder (PLADS). Even though those tests proved that airborne LASER bathymetry was feasible they were not conclusi ve in terms of operational system performance and cost benefit ratio. In 1973 a joint project between NAS A and the Naval Oceanographic Office was established to deve lop and thoroughly test an Airb orne LASER Bathymeter (ALB) System. The construction of the system was concluded in 1975. Using two pulsed LASER sources: a 2 kW 540 nm 6 nsec 10-200 pps Ne on Ion LASER and a 2MW 532 nm 8 nsec 10-50 pps frequency doubled Nd Yag LASER. The receive r consisted of a 28 cm Cassegrain telescope, two narrowband 0.4 nm filters and an 8575 RCA photomultiplier tube. Over 200 hours of fight test were used to collect bathymetric data on Chincoteaque, VA, the Chesapeake Bay and Key West, FL. (Kim, Cervenka, Lankford; 1975) In 1975 NASA Wallops Flight Center and AVCO Everett Research Laboratory proposed a bathymetric LIDAR Airborne syst em building on the previous experiences of the ALB. The proposal passed in 1977, and the new instrument was built in the same year by the AVCO Everette Corporation. The system became known as Airborne Oceanographic LIDAR (AOL). The program was sponsored by NASA, NAVOCE ANO, and NOAA as an advanced testbed

PAGE 27

27 sensor for altimetry, bathymetry, hydrography, and fluorosensing resear ch. (Guenther et al., 1978; Mitchell, N.D.). The first implementation of AOL was ba sed on AVCO C-5000 ga s (neon/nitrogen) 540.1 nm 2 kW 400 pps LASER, a 56 cm nutating scanner mirror, a 30.5 cm diameter Cassegranian f/4 telescope, a narrow band 0.4nm filter and a photomultiplier tube. This was the first implementation of a scanning airborne LI DAR altimeter. (Guenther et. al., 1978) The AOL was improved and redesigned several times over two decades. Improvements in enabling technologies such as differential GPS, Inerti al Measuring Units (IMU) and best estimate trajectory algorithms continually enhanced the precision, accuracy and productivity of the AOL system. In 1994 the AOL underwent major hard ware renovation and its topographic mapping and ocean fluorosensing functionalities were sepa rated in to two new systems: The Airborne Topographic Mapper (ATM) and the AOL. Figure 1-5. NASA AOL system. A) P3 Orion platform. B) Scannig mirror. C) Instrument rack. (Source http://sealevel2.jpl.nasa.gov/jr_oceanographer/oceanographer-williams.html last accessed March 20, 2007) The key technological advances that enabled the simple LASER profilers to become scanners were the development of more sensitive photodetectector s in the form of photomulipliers and avalanche photo diodes (which enabled the recovery of the reflected LASER pulses at lower signal-to-noise -ratios), LASER beam steering mechanisms (scanners and their respective encoders), and finally the the highly accurate Inte grated Navigation Systems (INS).

PAGE 28

28 The greatest advantage of airborne LASER s canning over original profiling methods is the dramatic increase in mapping cove rage per unit of flight time. Before 1995 all airborne LIDAR systems were custom built, highly expensive, and were only available for big research institutions and companies. Ho wever in 1995 Optech Inc., a company based in Toronto, Canada, offered the first commercial-off-the-shelf Airborne LIDAR Mapping System. Soon after other companies started offering their own systems: Riegl in 1996 with its low altitude scanne r, Saab Survey Systems in 1997, and the Top Eye and Azimuth Corporation in 1998 with its Aeroscan sy stem (which later was bought by Leica) (Baltsavias,1999; Flood,2001; Cheves, 2002). As technology advanced the receiving electr onics increased in sensitivity and timing accuracy, LASERs became smaller yet more powerful and robust, scanners became smaller and their encoders more precise. The continued mi niaturization and improvements of the airborne LASER scanners led the manufacturers to consider the construction of small, short range LASER scanners that could be mounted on tripods or sma ll vehicles. The first units were introduced in the late 1990’s. Some referred to these units as terrestrial LASER scanners to differentiate them from their airborne counterparts. Today there is no universally accepted nomenclature as each manufacturer has created its ow n terminology. Riegl names them “3D Imaging Sensor”, Optech calls them “LASER Ranging and Imaging Syst ems”, Leica labels them “High-Definition Surveying Systems”, but in the government and a cademic literature they are mainly referred as terrestrial LIDAR scanners. Over the last 8 years Terrestrial LASER S canners (TLS) have become very common and useful tools, mainly used in the areas of architecture, civil engineering, surveying and mapping. As mentioned earlier, science is always in need of measurements to model the world. Just as

PAGE 29

29 their predecessor, the Airborne LASER Scanner, proved to be useful to different braches of natural science, TLS systems have great potential to provide de nse and accurate range sampling of the environment in an efficient fashion. 1.3 TLS Subsystems and Principles of Operation Terrestrial LASER Scanners are complex and ve ry precise instruments; they are composed of two basic subsystems. The first subsyste m is a LASER ranging device commonly called LIDAR for (Light Detection And Ranging) or LADAR (LASER Detection And Ranging). The latter is the nomenclature generally used by th e military. The second subsystem is an optical and or mechanical device capable of steering the LASE R beam in a scanning fashion over the area of interest. (Frhlich & Mettenlei ter, 2004) Even though TLS have these common elements, there are at least a dozen different types of instruments in the current market. This is due to the fact that there are 3 different ways a LIDAR can work and more than 3 ways to do scanning in two dimensions. 1.3.1 LIDAR Ranging Principles A LIDAR can be used to measure the distan ce between two points in any of the three following ways: phase difference, time-of-flight and by optical triangulation. Each of these ranging approaches has its own se t of strengths and weaknesses. In the phase difference units, the continuous LASER signal can be modulated at very high frequencies, and the numbers of ranges per second is generally limited only by the sp eed at which the data can be recorded. Thus PD units have the largest point collection throughput of all type s of units. Their main limitation is the ambiguity of the range measurement. PD units are ideal in situations where very short acquisition times are desired, with sample collection rates in th e order of hundreds of thousands of samples per second and mm level accuracy, but with limited range.

PAGE 30

30 Optical Triangulation units ar e ideal for measuring distan ces of a few meters with micrometer level accuracy at high data rates. However, its accuracy depends on the relation between range and baseline distance and falls o ff rapidly with increasi ng range. Other limitations are that its performance can be degraded if the surface is not uniform in shape or reflectance and by the presence of noise in the form of ex terior illumination fr om non-coherent light sources.(Curless & Levoy;1995) Time of Flight units have the advantage th at they provide unambi guous ranges from a few meters all the way to thousands of meters. Ho wever, because after th ey emit the LASER pulse they must wait until there is a return signal be fore they can send the next pulse. The point collection throughputs are re latively low compared to the PD and OT units. However, is worth mentioning that there are some special airborne and long range LIDAR systems that are able to work with multiple pulses at a given time. The other limitation is that the range resolution (the ability to separate proximate obj ect at different range s) decreases as the pulse width increases. That means that for pulsed LIDAR there is alwa ys a tradeoff between range and range resolution. However they are established as the most common type of ranging LIDARs. 1.3.1.1 Phase difference measurement (PD) Phase Difference was the ranging method used in early geodetic EDMs like the Geodimeter & Tellurometer and on current syst ems that employ continuous wave (CW) LASERs. In these LIDARs the amplitude of th e LASER “wave” is modulated and the phase difference between the outgoing and reflected wave is measured. The problem with this method is that phase differences are not unique, ther e is always an ambiguity about the number of complete modulating wave cycles that have occu rred prior to the phase difference. Most current TLS that work under this modality do not provide ambiguity resolution so they have a limited range usually less than 100 meters (Wehr & Lohr 1999 ;Frhlich & Metten leiter, 2004). It is

PAGE 31

31 important to clarify that the LASER has its own natural frequency a nd wavelength. Optical LASERs have wavelengths in the range of .4 to 1.5 m. The phase difference that is used to determine the target range is based on the am plitude modulating signal (not the LASER natural wavelength), which has a wavelength in the order of several to hundreds of meters. Figure 1-6. Phase difference ranging principle. Table 1-1. Phase difference ranging pr inciples expressions and equations. Parameter Formula Ambiguous phase difference n APD 2 Unambiguous phase difference before one full wave cycle 2 way travel time before one full wave cycle c f tlong carrier way 2 22 Long amplitude modulating signal wavelength carrier longf c Short amplitude modulating signal wavelength short Speed of light c Signal to noise ratio s/n Maximum unambiguous range 2max longR Range 4 2 2 1long carrierf c R Range resolution 4shortR Single shot range accuracy n sshort R/ 1 4 ~

PAGE 32

32 1.3.1.2 Time of flight (TOF) In PD LIDARs the phase difference of the continuous LASER is used to determine the 2 way time-of-flight of the modulated signal. As mentioned in the previous section, the main disadvantage of this method is the ambiguous ra nge. An alternative that help overcome this limitation came with the development of Q-switching by McClung a nd Hellwarth in 1961 (McClung & Hellwarth,1962). This invention enabled the emissi on of very energetic LASER pulses rather than the continuous wave beams. Th is allowed to directly measure the 2 way timeof-flight without any ambiguity. Ho wever, even when these pulses are relatively short in time, generally in the order of a few na noseconds, at the high speed that light travels this translates into several centimeters (e.g. 10ns = 3.0 m). In or der to obtain the sub-centimeter accuracy an specialized electroni c circuit called a Constant Fraction Disc riminator (CFD) is used to precisely time a specific point on the pulse (generally the half point of the pulse amplitude). With discrete packets of light and the CFD is a simple ma tter to measure with high accuracy the time difference between the emission of the pulse and the detection of its reflected return. This method provides unambiguous range measurements of distances limited only by the dispersion of the LASER energy and the sensitivity of the detector (Wehr & Lohr, 1999). Figure 1-7. Time of fli ght ranging principle.

PAGE 33

33 Table 1-2. Time of flight ranging pr inciples expressions and equations. Parameter Formula 2 way travel time t Pulse width t Speed of light c Signal to noise ratio s/n Pulse rise time pulseriset Range t c R 2 1 Range resolution t c R 2 1 Single shot range accuracy n s t crise pulse R/ 1 2 ~ 1.3.1.3 Optical triangulation (OT) In PD and TOF the outgoing and incoming LASER beam follow the same optical path, however on OT the reflected waveform is observed from a different vantag e point. In OT units a LASER beam is steered by a scanning mirror over the target surface and its reflection is collected through a lens that focuses an image on a position sensitive detect or such a CCD array. The position of the spot image on the pixels of the camera, the scanner angle and the LASER to lens optic center baseline is then processed us ing trigonometry to determine the distance to the target. (Beraldin et al., 2003) Figure 1-8. Optical triangul ation ranging principle.

PAGE 34

34 Table 1-3. Optical triangulation rangi ng principles expressions and equations. Parameter Formula Baseline distance D Deflection angle of the LASER beam Position of the imaged spot p Focal length f Range tan f p f D Z “Horizontal” position tan Z X Position accuracy p Single shot ranging accuracy p zD f Z 2~ 1.3.2 Scanning Mechanisms Traditionally, geodetic scanning instruments have used reflective optics coupled to a mechanical system, although some newer instrume nts (CATS & Jigsaw) use refractive scanning elements such as Risley prisms (Carter et al., 2005; Marino & Da vis, 2005). Current TLS systems have a 2 axis capability. This is achie ved in its simplest implementation by a single line scanner through a rotating or nut ating mirror, and the second scanning axis is obtained by rotating the complete instrume nt as shown in Figure 1-9. Figure 1-9. Scanning axes on a panoramic view s canner. (Adapted from original source at: http://www.riegl.com/terrestri al_scanners/lms-z390_/390_all.htm last accessed March 20, 2007)

PAGE 35

35 This type of scanner is usually called a Panoramic View Scanner and is capable of scanning 360 in azimuth and from +75 to -75 in elevation. The other type of scanner is called Camera View Scanner. It is usually implemen ted by two perpendicular deflection mirrors, one for the azimuth and the other for the elevation. Th is type of scanner has a fixed field of view typically of 45x45, but it can be extended to a pa noramic field of view with the aid of optional pan and tilt bases. (Frh lich & Mettenleiter, 2004) 1.4 Technical Characteristics and Specifications of TLS Currently there are more than 12 commercial-off-the-shelf TLS systems. With this wide range of options, defining which is the right TLS for a project depends on the careful analysis of the instruments specifications and the project requirements. Appendix A has a description of commercial TLS units. Key specifications of TLS are: range, range resolution, precision, accuracy, azimuth & elevation resolution or poin t spacing, LASER type & wavelength, and the scan rate, field of view. Subsystem specific characte ristics will be described first as they have an impact on integrated system characteristics. LASER wavelength: Even when the manufactur er won’t provide the exact wavelength of the LASER it will generally give a range such as green, red or near infrared. Some consideration must be given to the LASER wavelength of the TLS Visible LASERS will be best when there is the need for water and glass penetration or wh en mapping wet surfaces. Mo st of the energy of infrared LASERs will be absorbed by moist su rfaces and the return si gnal will be very weak. 1.4.1 LASER Type LASER type will be either pulsed or conti nuous wave. The advantage or disadvantage of each type was discussed in the previous section.

PAGE 36

36 1.4.2 LASER Class The Class of a LASER is defined by the American National Standards Institute (ANSI) according to the degree of hazar d presented to eye safety based on a maximum permissible exposure (MPE). The class depends on the LASER power and wavelength. TLS must be built to meet eyes safety regulations, and the operator must be aware of what are previsions and precautions that must be taken. Table 1-4 pr ovides a summary of the LASER classification scheme. Table 1-4. LASER classification Class Description Class I Safe for use under all reasonably-anticipated c onditions of use. It is not expected that the MPE can be exceeded. This class may include LASERs of a higher class whose beams are confined within a suitable enclosure so th at access to LASER radiation is physically prevented. Class IM This LASERs produce large-diameter beams, or beams that are divergent. The MPE for a Class 1M LASER cannot normally be exceeded unless focusing or imaging optics are used to narrow down the beam. If the beam is refocused, the Class has to be upgraded. Class II Emits in the visible region. It is presumed that the human blink reflex will be sufficient to prevent damaging exposure Class IIM They emit in the visible region in the form of a large diameter or divergent beam. It is presumed that the human blink reflex will be sufficient to prevent damaging exposure. Class IIIR All continuous wave LASERs which ma y produce up to five times the emission limit for Class 1 or class 2 LASERs. Although the MPE can be exceeded, the risk of injury is low. The LASER can produce no more than 5 mW in the visible region. Class IIIB They produce light of an intensity such that the MPE for eye exposure may be exceeded and direct viewing of the beam is potentially serious. Class IV High power (typically more than 500 mW if cw, or 10 J/cm if pulsed). Th ese are hazardous to view at all times, may cause devastating and permanent eye damage, may have sufficient energy to ignite materials, and may cause significant skin damage. 1.4.3 LASER Beam Divergence The LASER beam divergence will determine the footprint area at a given range. For an accurate mapping one will require the smallest f ootprint size because th e LIDAR computes an average range of the entire illuminated area. Th e larger the area, the more chance of slope, reflectivity and smoothness variations affecting the range measurement.

PAGE 37

37 1.4.4 Point Spacing Point Spacing or angular resolution (azimuth a nd elevation) is the measure of the smallest angular step the scanner mechanism can steer the LASER beam. In other words, it is the measure of the angular or linear separati on between adjacent LASER shots. 1.4.5 Range Range is perhaps the most important ch aracteristic of a TLS, and performance specifications should include both a maximum a nd minimum ranging distance. Range will vary greatly between units based on the LIDAR prin ciple of operation and the specific design characteristics (pulsed energy a nd detector sensitivity for TO F, long wavelength for PD and baseline distance for OT). 1.4.6 Range Resolution Range resolution refers to the ability of the TLS to distinguish between adjacent features in the range direction. Range reso lution depends on pulse width on TOF units, of phase measuring resolution on PD, and on baseline distance and the spatial resolution of th e position detector in OT units. 1.4.7 Precision and Accuracy Precision and accuracy are often used usually interchangeably, however, they are not the same concept. Precision is the statistical clos eness of a set of repeated measurements while accuracy is closeness of the best estimate va lue obtained by the measurements to the accepted true “value” of the measured quantity. (DMA, 1991) Baltsavias(1999) states that “ranging precision” is inversely proportiona l to the square root of the signal to noise ratio and gives expressions for the estimate of the “ranging pr ecision” for TOF and PD LIDARs. Wehr & Lohr (1999) present the same expressions but they refer to them as “r anging accuracy”. What they are in fact are referring is not to either ranging accuracy or precision but rather estimates of ranging

PAGE 38

38 errors as a function of intrinsic electronic parameters which will have some relation to the ranging precision and accuracy. Before describing TLS precision or accuracy it is necessary to consider that they reconstruct reality by measuring ranges and angl es. The final coordinate s of points are derived from computations using the observed ranges and angles, and they are subject to a combination of errors from each of the hardware subsyste ms as well as rounding and other computational errors of the software subsystem. It impossibl e to derive closed form equations for a TLS precision and accuracy; this parameters must be estimated through extensive laboratory testing. The overall precision of a TLS system is th e degree of repeatability of its range and position measurements. There are two types of precision, single meas urement precision and averaged measurement precision. The single me asurement precision can be understood as the theoretical error in measuring a single point only once. Averaged measurement precision is obtained if the system takes multiple measurements of a single point and computes a mean and standard deviation from these observations. To determine accuracy a manufacturer has to test the measurements derived from TLS us ing a higher quality data set to see how well they agree. Manufactures usually quote the modeled accuracy, which is derived from the fitting multiple point measurements to a primitive model. (Iavarone, 2002)

PAGE 39

39 CHAPTER 2 UF MOBILE TERRESTRIAL LASER SCANNING (M-TLS) SYSTEM 2.1 Evolution of the M-TLS Concept The University of Florida (UF) was a pioneer in the application of commercial airborne laser mapping systems to the fields of environmen tal and infrastructure surveying. In October, 1996, a demonstration/test project was conducted for the Flor ida Department of Environmental Protection in collaboration with Optech, the Fl orida Department of Transportation (FDOT) and the US Geological Survey Center for Coastal Geology. During project LASER (Laser Swathmapping Evaluation and Resurvey) more than th ree hundred kilometers of beaches (Mexico Beach, FL, to the western tip of Perdido Key, AL) and a portion of Interstate 10 were mapped using an Optech Inc. ALTM 1020 laser ranging sy stem. (Carter & Shrestha, 1997). Shortly after, in 1998 UF in conjunction with the Florida Inte rnational University (F IU) purchased its first LIDAR mapping unit, an Optech 1020 ALTM. In early 2001, the University of Florida (UF), the National Geodetic Survey (NGS), the FAA, and Optech, Inc. conducted a field test to explore using an Airborne Laser Scanner for Detecting Airport Obstructions at the Gainesville Regional Airport. (Parrish et al., 2005) During this research UF personnel tested a preproduc tion prototype of an Optech’s tripod mounted LIDAR unit named “Intelligent Laser Ranging Imaging System” (ILRIS-3D). After the tragic attacks on September 11, 2001, the Office of the Deputy Under-secretary of Defense approached the US Army's Joint Pr ecision Strike Demonstrat ion (JPSD) to inquire about specific technology cap abilities to aid the survey ing the NY ground zero. JPSD approached Optech and the University of Flor ida Geosensing Systems En gineering center for personnel and equipment support. A multi inst itutional group was established and laser

PAGE 40

40 surveying began on September 18. Besides the airborne unit, two ILRIS-3D systems were deployed to map rubble piles and damage d building structures (Kern, 2001). These experiences with the Terrestrial Laser Scanner made the UF researches quickly realize the potential of the technology as a va luable complement to the airborne laser unit. During 2002 the UF Geosensing Systems Engineeri ng (GSE) division acquire d and fully tested an ILRIS-3D unit. The results of the tests ar e thoroughly discussed in the 2002 Master thesis “Applications of Laser Scanning and Imaging Systems” by GSE student Devin Robert Drake. In 2002 UF prepared a proposal for the Florida Department of Transportation of a Mobile Laser Surveying System (MOBLASS). Potential applications of the system were identified to include the precise positioning, the continued evaluation and documentation of the primary components of the transportation infrastruct ure (center lines, guardrails, signs, bridges maintenance facilities). The proposal consisted on a twelve passenger van equipped with an ILRIS 3D unit mounted on a telescopic pan tilt base, a GPS array for positional and azimuth determination, power leveling and stabilizing units an auxiliary power generator, an operator control console with display, a PC for data reduction and analysis, a wireless data link for realtime transmission to the ope rational center and upgrade capabil ities for an IMU and other types of sensors (digital imaging, hyperspectral, ground penetrating radar). At that point in time technology just permitted static mapping and the proposal was developed as a Stop-Map-and – Go system, unfortunately FDOT decided not to fund the development of the MOBLASS. Figure 2-1 shows a survey vehicle developed around a concept similar to MOBLASS. In 2005 the UF Geosensing Systems Engin eering division decided to develop the MOBLASS concept with its own resources. Th e system name was changed to M-TLS for Mobile Terrestrial Laser Scanning system and is based on a new version of the ILRIS 3D system

PAGE 41

41 which enables the interfacing with an Optech pr oduced Pan Tilt base. Other mayor change in the concept is that the vehicle is now a 4x4 truck which extends the terrain operation capabilities and allows an increased instrumental payload. Th e M-TLS is a unique tool that enables the acquisition of high density point clouds from an advantageous terrestrial geometry, being a very valuable complement for Airbor ne Laser Scanner data sets. Figure 2-1. Survey vehicle with a similar concept to MOBLASS. 2.2 M-TLS Subsystems 2.2.1 LIDAR Unit The core of the M-TLS is a commercial 2-axis ground based laser scanner which is integrated to a mobile telescoping, rotating and ti lting platform which provide up to 6 degrees of freedom for performing scanning op erations. A scanner built-in 6 me gapixel digital camera and a digital video camera provide the M-TLS still and video imagining capability. The laser scanner is an Optech ILRIS-36D, which is capable of generating XYZ with laser intensity or RGB textured point clouds in a ra nge from 3 m to 1500 m for a target with an 80% reflectivity or 3 m to 350 m to targets with a 4% reflectivity. The laser op erates at a wavelength of 1535 nm, with a pulse width less th an 10 ns and energy of less than 10 joules. The sample separation can be adjusted down to 0.00115, and the scanning speed is 2,000 points per second.

PAGE 42

42 Currently the ILRIS is only capable of pe rforming its scanning and mapping operations on a static mode. In a near future Optech will release a unit with “on-the-move” mapping capabilities, at that point is expected that the M-TLS will perform dynamic mapping operations. Figure 2-2. OPTECH IL RIS block diagram. Figure 2-2 shows a block diagram of the ILRI S unit. At a very high level the internal operation of the ILRIS is quite simple. A cen tral microprocessor controls the different subsystems, collect, analyzes and displays data and information. The computer commands a laser controller to fire a pulse, the la ser beam generated by a Nd YAG laser is passed thru a non linear crystal that shifts its natural frequency from 1064 nm to 1535 nm. Then the laser beam pass thru a beam expander and a small amount of photons are di verted thru a fiber op tic cable to start the time of flight timer. After the beam is expanded it passes through an optical element that reflects the returning beam to the detector. The beam is then deflect ed by the vertical and horizontal scanning mirror to the target. The position of each of the mirrors is controlled by the scanner axis drivers and the

PAGE 43

43 central computer. Most of the targets have quasi -Lambertian reflective properties and will reflect the laser beam as a distorted waveform, part of that waveform will return in the same optical path of the outgoing beam. The incoming waveform will be reflected by the scanning mirrors to a fixed parabolic mirror that focus th e waveform on to the detector. Prior to entering the detector a narrowband optical filter centere d at the 1535 nm limits the noise entering the system. When the returning laser photons arrive at the detector, which is an indium gallium arsenide (InGaAs) Avalanche Photo Diode (APD), a voltage is gene rated between its terminals. The output voltage is read by an A/D converter wh ere the signal is digitized and sent to a Constant Fraction Discriminator (CFD). The CFD is an electr onic device designed to produce accurate timing information from signals of varying amplitudes bu t the same rise time. CFD usually achieve this by splitting the input signal, attenua ting half of it and delaying the other half, then feeding the two halves into a fast comparator with the delayed input inverted. By doing this CFD is capable of triggering a timing signal at a constant frac tion of the input amplitude. The CFD trigger is fed into the precise Time Interval Meter (TIM) which was original started by the outgoing pulse feed to the detector thru the optical fiber. The TIM computed the time difference between the outgoing and incoming pulse thus determining th e 2 way time-of-flight. The computer then calculates the range to the target records it on it internal memory and commands the emission of a new pulse and the entire process is repeated until the defined ar ea of interest is scanned. The block diagram shows additional subcom ponents of the ILRIS unit. On-board 6megapixel CMOS digital camera is used to provid e the operator with a lo w scan rate video and high resolution stills of the scan area. The Came ra is boresight calibrated with the LIDAR to provide red, green and blue channels to each la ser point. The output from the camera is also projected with other control information to a 5. 5 ” x 4 ” LCD viewfinder. On board data storage

PAGE 44

44 is done on a conventional USB jump drive. The controlling and commanding of the ILRIS is performed thru proprietary software called “Cont roller” which runs from a computer or pocket pc. To provide connectivity to the PC the ILRIS has two network interfaces devices: a Ethernet interface card (IEEE 802.3) and a wireless netw ork card (IEEE 802.11). Power to the ILRIS is provided from a 28 V DC battery pack or from a 120 VAC/28 VDC power adapter. 2.2.2 Vehicle The vehicle selected for the M-TLS system is a Ford F250 4x4 crew cab long bed truck. The truck has undergone several modifications in cluding the installation of a steel frame for mounting the telescopic lift and the installation of four electrical jacks which will enable automatic vehicle stabilization and leveling. Th e vehicle 4x4 capability al lows the execution of off-road mapping projects, its 1600+ kg of cargo capacity allows the loading of the 374 kg lift with ample capacity for mapping, positioning and auxiliary equipment. Figure 2-3 shows the MTLS truck with the lift deployed at half height. Figure 2-3. Mobile Terrestrial Laser Scanni ng (M-TLS) system: truck, lift and ILRIS.

PAGE 45

45 2.2.3 Lift The lift is an AC powered JLG Push-Around Ve rtical Lift model AM 25. This lift has a stowed height of 1.97 m and once deployed it extends to 9.45 m. The truck was modified by bolting a steel frame directly to the back chassis, the lift then rests over the steel frame. An instrument aluminum frame was constructed to support the ILRIS, its power supply and an electronic tilt meter. The instrument frame is mounted on the top of the lift. With all these provision the ILRIS scanner can be lifted to a vantage point of up to 12 meters above ground level. 2.2.4 Power Subsystem The M-TLS have components that require bo th AC and DC power. To supply the power requirements the M-TLS system has a Briggs & Stratton 6200 Watts electri c start generator and a 12V DC battery bank. The electr ic generator is used when the lift needs to raised or lowered and to charge the battery bank. DC power is di rectly available from the battery bank and low power AC devices can be fed thru a DC to AC power inverter from the battery bank. Figure 2-4 shows the power plant installed on the truck bed. Figure 2-4. M-TLS deployed in Ge orgia, showing the power genera tor and field computer on the truck bed.

PAGE 46

46 2.2.5 Pan Tilt Base Currently there are two pan & t ilt bases available on the M-TLS system. The first one is an Optech manufactured base that is connected to the ILRIS 3D unit and controlled from the ILRIS control software. This base allows a complete 360 rotation in azimuth and 70 in elevation, with the 40x40 ILRIS’s field of view the tilt base permits a -20 to 90or a -90 to -20 elevation coverage. The advantage of using this base is th at the scan data can be automatically de-rotated and de-tilted in the Parsing proce ss, yielding a complete coherent data set. The disadvantage is that when the ILRIS is powered up and it detect s that is connected to a pan tilt base it will capture a 360 panoramic picture composed of 10 individual digital camer a frames and this can be extremely time consuming if the user is ju st interested in scanning a narrow field. To overcome this disadvantage a second pan tilt base is available; the base was manufactured by QuickSet International and is desi gned for the operation of survei llance cameras. It has a loading capacity of 40 kg and a rotation range of 217.5 in azimuth and 90 in elevation. This pan tilt base can be controlled from a PC or from an anal og joystick console. Figure 2-5 shows the two available pan tilt bases for the M-TLS. Figure 2-5. M-TLS pan tilt bases. A) Optech 360 base. B) Quickset QPT pan tilt base.

PAGE 47

47 2.2.6 Video Camera The MTLS is equipped with a Samsung SCCC4201 high resolution color video camera. The camera has a built in X22 optical lens and an X10 digital zoom to provide an effective, enhanced focal length of 3.6 to 79.2mm. Th e 410,000 pixels CCD ( 811 x 508) outputs NTSC with 480 Horizontal Lines and 350 Vertical Lines. The video pr ovides the operator with wide and narrow views of the scan project area. The out put can be directly viewed on a TV screen or on the computer thru a video capture card. The vi deo camera can also be used as a surveying and mapping tool during operation of the M-TLS dyna mic mode. The video camera is mounted on top of the ILRIS unit by means of a speci al housing as shown in Figure 2-6. 2.2.7 On Board PC Currently only data collecti on is done onboard. For this pu rpose a laptop containing the ILRIS controller software is used. The ILRIS to PC connection is done by the wireless peer to peer network. Data is stored in both the ILRI S USB disk and on the laptop hard drive. On a future a central computer that will control and re cord the data from the ILRIS, INS and cameras will be installed. 2.2.8 GPS At this point in time the ILRIS unit can not do direct georeferencing. In the static mode, when a geo-referenced dataset is desired the us e of GPS control points is required. This is usually done by surveying at leas t three control points well di stributed both vertically and horizontally on the area to be scanned. An additi onal control point can be located by installing a GPS antenna on top of the ILRIS unit, by adding the proper XYZ offsets the GPS coordinates of that antenna reference point can be translated to the ILRIS coordinate origin. Figure 2-6 shows the provisions for the installati on of a medium size L1/L2 marine antenna on the ILRIS unit. For the purpose of surveying the control points a se t of geodetic grade GPS receiver are available

PAGE 48

48 which include units from the ASHTECH models Z-Xtreme, Z-Surveyor and Z-12. Ashtech L1/L2 Choke Ring and marine L1/L2 model 700700 ar e used to collect the GPS signals. Figure 2-7 shows an array of GPS stations used for geo-referencing of M-TLS data set during the forestry experiment. Figure 2-6. The ILRIS unit with on ax is video camera and GPS antenna. Figure 2-7. Array of geodetic quality GPS base stations used for geo-referencing.

PAGE 49

49 2.2.9 Tiltmeter An electronic tilt meter is installed under the instrument frame on the lift to measure the vibrations and motions to which the ILRIS is subjected and to ensure that during scanning operations the platform remains as steady as pos sible. The red ellipse on Figure 2-8 marks the installation position of the Tiltmeter unit. Figure 2-8. Installation of the tiltm eter unit on the instrument frame. 2.2.10 INS When Optech releases the new generation of ILRIS with on the move scanning capability, the M-TLS will be capable of performing mapping ope rations in the dynamic mode. At this point the installation of an Integrated Navigation System must be performed. This can be in the form of a canned solution like the Applanix POS LV or with an in-house developed INS. The Applanix POS LV system includes a single freq uency two antenna survey grade GPS array and an Inertial Measuring Unit. The IMU grade de pends of the clearance obtained by the US Department of State regarding the Inte rnational Traffic in Arms Regulations.

PAGE 50

50 CHAPTER 3 WORKFLOW OF M-TLS OPERATIONS 3.1 Data Collection Data collection design is the first step in M-TLS operations. There are no written rules on how to perform a data collection, it depends on very specific details of the project and the experience of the operator. Data co llection design starts with an analysis of the requirements of area of interest; desired resolution or laser point spacing; laser return intensity or RGB texture; data set reference (sensor XYZ or geo-referenced to a particular datum) and accessibility to the scan area. With these inputs the op erator defines the data set acqui sition strategy that includes the selection of the scanning geomet ries (number and orientation of the scans) and the design of the GPS Control Points Network if the data set needs to be referenced to a particular datum rather than the natural XYZ sensor frame of reference. Collection is performed with th e Optech proprietary software “Controller”. The controller main screen displays a color image which c overs the 40x40 scannable field of view. The operator then selects a Region of Interest (ROI) to be scanned and based on a preliminary range acquisition, adjusts the spot spac ing in angular or linear units. After setting the data storage directory the scan can be initiated. When us ing the Optech pan and tilt base the ILRIS will acquire a 360x40 panorama consisting of 10 ove rlapping still frames. The complete panorama will be displayed in Controller window and the us er can then set the ROI to be scanned. Figure 3-1 presents a screen capture of the Controller software. For particular projects several setups are required. In some cases a combination of panoramic and frame scans have to be performed. The integrated M-TLS system provides up to 6 degrees of freedom (4 rotations and 2 translat ions) for performing scanning operations that guarantee that all possible facets of the obj ect will be mapped under the most favorable

PAGE 51

51 conditions. Figure 3-2 shows photogra phs of different configurati ons of the M-TLS used during data collections. Figure 3-1. Screen capture of the ILRIS unit controller software dur ing scanning operations. Figure 3-2. Different configuratio ns of the M-TLS system during se veral data colle ction projects. A) Forest scanning B) Soil roughness expe riment, a microwave radiometer can be seen behind the M-TLS. C) Beach erosi on hot spot mapping in St. Augustine Fl. 3.2 Data Parsing At collection the data is stored in a bina ry format and includes sensor orientation parameters and range for each measurements made. The raw data needs to be converted into a

PAGE 52

52 position in a coherent 3D frame of reference. Additional information can be texture or color information from the intensity of the reflected la ser signal or from a coregistered imaging sensor. This first step of processing is called Parsing and it is performed by the proprietary software “PARSER”. With Parser the raw data is convert ed to any of the known point cloud formats that can be read by most LIDAR processing softwa re for further manipulation and analysis. Figure 3-3 shows a screen capture of the Pa rser software. The large screen shows the digital image of the scan surface, the red box enclos es the selected region of interest. The Parser Settings tab allows the user to personalize th e conversion process. There are options for the output file format, the choice of using the digita l image to provide RGB color channels for each of the scanned points, smoothing of the range shot s, corrections based on the Pan-Tilt orientation and an option for the positioning of the measurement origin. Figure 3-3. Screen capture of the Parser soft ware during the setup of parsing settings.

PAGE 53

53 The most common output select ed is the ASCII XYZ File which converts the range, azimuth and elevation into a coordinate in a right hand Cartesian frame with the origin at the LIDAR reference point. Additional information can include a normalized 8 bit laser return intensity or RGB channels. This format is easily uploaded into point clou d processing software like QT Modeler, Terra Scan, Polyworks or Matlab for a customized analysis. 3.3 Data Manipulation and Information Extraction Once the data is parsed is ready for uploadi ng into specialized software that allow its manipulation, analysis and information extractio n. The information extraction from a M-TLS is the final goal, it is a customi zed process for each application. Some of these specialized techniques are discussed on an appl ication to application basis on ch apters 4 to 7. However, there are several typical operations that can be performed on LI DAR point cloud data; these are visualization, transformations, segmentation, clas sification, filtering, gridding and specialized mathematical operations. 3.3.1 Visualization The first thing that is done to a LIDAR data se t is to look at it. Visu alization is the most basic operation; however, a good visua lization allows the analyst to asses the quality of the data set, it enables the planning and control of diffe rent processing schemes a nd finally will provide the presentation of the final product. Most LIDAR processing software will have a graphical interface that will render the numerical point cloud into an image, but ther e is a great range of options and functionalities that will vary among the different opti ons. The simplest visualization will plot all the points with a single color and size, and the operations of Zoom, Rotate, and Navigate will be available. More advanced soft ware will render each LIDAR point according to other characteristics, it can be brightness coded according to the laser return intensity, or RGB textured if the point cloud was coregistered with a digital image, it can also be color coded

PAGE 54

54 according to elevation, range, class or any other attribute contained in th e point cloud structure. Some software will allow the user to toggle between the rendering of the point cloud and the rendering of a Triangulated Irregul ar Network (TIN) or Digital El evation Model generated from it, as shown in Figure 3-4. Figure 3-4. Visualization of LIDAR data. A) As a point cloud. B) As a DEM. 3.3.2 Single Point Selection An important functionality of visualization software is the one that allows the user to manually do single point selection. This is to navigate through the poi nt cloud using the zoom and rotate controls to pick out single points from the cloud. 3.3.3 Measurements The ability to precisely select points from the clouds allows the analyst to make measurements such as distances between points, and angles between lines connecting the points. 3.3.4 Primitive Fitting After selecting a series of points is possi ble to perform a primitive fitting operation. Primitive fitting is the applic ation of the least square methodology to compute the spatial parameters that define simple geometric figures or volumes such as lines, circles, planes, spheres, cones. Primitive fitting allows computing the modeling accuracy of TLS. Imagine that

PAGE 55

55 there is a sphere with a known radius. After the sphere is scanned a best fitting sphere is determined for the point cloud using least squa res to minimize the residuals. The “fitted” or “modeled” sphere radius can then be compared to the known radius and the degree of agreement between both provides an estimate of the modeli ng accuracy. Figure 3-5 i llustrate the entire process from point selection to the fitting of the sphere. Figure 3-5. Primitive fitting process illustrated. A) Point cloud. B) Picking points from the sphere surface. C) Fitted sphere based on the picked points. 3.3.5 Generating Cross Sections An important visualization tool is the selecti on of a particular baseline and the generation of a cross sectional view of th e point cloud at that baseline. Figure 3-6. Cross section gene ration from the point cloud.

PAGE 56

56 3.3.6 Transformations There are countless transformations that can be applied to the point clouds, a few of the most frequent are described following: 3.3.6.1 Rotations and translations Simple Transformation includes the translation or rotation of the en tire point cloud on one or more of the coordinate axes. 3.3.6.2 Cropping When an object is scanned there are always points that do not be long to the volume of interest. Cropping allows the creatio n of a point cloud with only the elements that falls within the 3D space of interest. 3.3.6.3 Merging Point Cloud Merging is performed when seve ral point clouds of the same object were collected from different angles or positions each having its own coordinate frame and there is the need to convert all of them into a single spa tial coherent point cloud. Merging is performed by setting one point cloud as the base referen ce frame, and then common points or common primitives are identified between the base and the source point cloud. From the common points a 3D rotation and translation transformation is com puted using least squares adjustment. Then the transformation is applied to the source point clou d to change its coordinate frame to the base reference frame. Figure 3-7 illustrates a mergi ng operation between two point clouds color coded as white and pink that were obtained from di fferent scan angles, using the common points method the pink point cloud was rotated to the white coordinate system to produce a single coherent data set. With the Polyworks IMspect and IMAlign so ftware the common points can be manually selected and then the transformation process is done automatically. A nother option is to use

PAGE 57

57 Polyworks IMspect to find the common point c oordinates in both the source and the target frames create text files with t hose coordinates and then use Terra Scan transformation modules to compute the parameters of the transformation and then manually apply it to the source point cloud. The advantage of this latter option is that TerraScan provid es a residual analysis of the transformation based on the common points and this give the user an idea of the quality of the transformation. Figure 3-7. Point cloud merging example. 3.3.6.4 Geo-referencing A transformation in which a point cloud with coordinates in arbitrary sensor space is converted into a geodetic coordina te frame is called geo-referencing. This operation has to be performed when absolute measuremen ts have to be made or when the terrestrial data set will be merged or compared to an airborne data set. Similar to a merging operation, in geo-referencing there has to be a minimum of 3 non-collinear po ints for which coordinates on both sensor and geodetic frames are known. Based on that set of c oordinates the parameters of a 3D rotation and translation transformation are computed. That tran sformation is then applied to the entire point cloud and as a result the data set is fixed to th e specific geodetic frame. The accuracy of the georeferencing depends primarily of the quality of the GPS observations, the vertical and horizontal strength of the control points network and the determination of the XYZ coordinates of the control points from the original point cloud. Geo-referencing can be performed using both Polyworks and TerraScan.

PAGE 58

58 3.3.7 Segmentation, Classification and Filtering Another important set of operations performe d over the point clouds are the ones that allow performing segmentation, classification and filtering of the points. 3.3.7.1 Segmentation Segmentation refers to the operation that will segment or segregate points into different groups based on characteristics w ithout knowledge of what they really are. An example of segmentation could be the separation of points, based on intensity values, into low intensity, medium intensity and high intensity. Under this segmentation scheme points in each group will not necessarily share common spatial characteristics. 3.3.7.2 Classification Classification implies the separa tion of points into different gr oups or classes defined by an intrinsic or natural characteristic. An example of classification is the separation of the points into vegetation, building or gr ound classes; each of these groups im plies the knowledge of its nature. 3.3.7.3 Filtering Filtering is the removal of a set of points from the clouds based on either a segmentation or classification scheme. An example of a segmentati on scheme based filter could be the removal of points that are below a certain height value, w ithout considering its natu re (i.e. ground or low vegetation). A classification filter could be one that removes vegetation from an urban scene on which only brick and glass is wanted. 3.3.8 Gridding A scanner point cloud by nature is an irregula rly space data set. The process of converting the point cloud into a regularly spaced data set by m eans of interpolation is called gridding. Gridding allows the analyst to observe subtle feat ures in the data set. There are many different gridding algorithms the more common are N earest Neighbor, Inverse Distance Weighting,

PAGE 59

59 Triangulation with Linear Interpol ation, and Kriging. The regular nature of the grid allows the analyst to perform many mathema tical operations such as areas and volumes computations, grid algebra, grid calculus, differen tiation, gradients, grid comparison, as well as image processing operations. Gridding can be performed using speci alized software such as Golden software “Surfer” or with built-in routines in Matlab. Fi gure 3-8 shows a regular grid surface model B) generated from the irregu lar spaced point cloud. Figure 3-8. Gridding operations, from point cloud to grid. A) Point cloud to B) Surface Model 3.3.9 Advanced Mathematical Operations The operations discussed so far are comm only performed by the LIDAR analyst using canned algorithms in commercial software pack ages. However, some applications require

PAGE 60

60 advanced or specialized techni ques that must be custom pr ogrammed in programming languages such as Visual C, Visual Basic or using higher level math tools su ch as Matlab or IDL. Examples of these advanced mathematical operations may include: Transformations from space to the frequency do main using the Discrete Fourier Transform or with the Discrete Wavelet Transform. The use of spin images to represent objects from a 3D dataset in a single 2D image. The application of advance image processi ng techniques and operations such as edge detection or morphological opera tions to a gridded dataset. Figure 3-9 shows an example of and advanced mathematical operation. Were a discrete 2D Fast Fourier Transform was used to extract the stronger periodic components of the real terrain and then a 3D surface was generated from those components. Figure 3-9. Examples of advanced mathemati cal operations in the pr ocessing of Airborne LIDAR data. Digital elevation models of A) Gabilan Mesa, Ca. and B) South Fork Eel river, Ca. C) and D) show visual re presentation of the mo st strongly periodic component of each landscape. (Perron, 2006)

PAGE 61

61 CHAPTER 4 TESTED APPLICATIONS OF M-TLS 4.1 Common Applications of TLS TLS units are marketed by the manufacturers mainly as surveying tools for engineering applications. These engineering applications in clude as-built-surveys, crime scene and traffic accident investigation, mine operations planning and supervision, transportation infrastructure mapping, bridge loading analysis, building dama ge assessment and urban modeling. The reason for this biased marketing is that TLS are expens ive instruments that are often affordable only by companies that do large projects with huge capita l investments, where the savings of time over classical mapping techniques are even greater than the cost of the TLS instrumentation. But even with that biased marketing over the last 5 years, some articles about scie ntific applications of TLS have been published in scientific journa ls or presented in pr ofessional meetings. One of the first presentations reporting the application of TLS technology to scientific research was that of the University of Te xas Bureau of Economic Geology “3-Dimensional Digital Outcrop Data Collection and Analysis Using Eye-safe Laser (LIDAR) Technology” presented on the 2002 convention of the American Association of Petroleu m Geologists (Bellian et al., 2002). In 2003 a paper published in the In ternational Society for Optical Engineering (SPIE) Optical Metrology for Arts and Multimedia journal entitled “High-resolution laser radar for 3D imaging in artwork cataloging, reproduction, and restoration” intr oduced the application of TLS for cultural heritage pres ervation. With respect to forest ry one of the first articles published was “Using airborne and ground based ranging LIDAR to measure canopy structure in Australian forests” published on the Canadi an Journal on Remote Sensing vol. 29, 2003 (Hopkinson et al., 2004).

PAGE 62

62 These papers were the groundbreakers of the scien tific applications of TLS. It is clear that TLS has a great potential to contribute to severa l fields of science that require precise spatial measurements. It is just a matter of making th e technology available to researchers and providing them with technical support on th e data processing and informati on extraction. Just as Airborne Laser Mapping proved to be a valuable tool, TLS w ill be accepted if it proves to be better that conventional data gathering techniques by providi ng higher quality and quantity of data faster and cheaper. Several applications were tested under this philosophy, a subset of which are briefly described in the following sections. These sec tions present a “Traditional” vs “Alternative” methods of collecting spa tial field measurements. Subsequently, three of these applications will be thoroughly discussed in the next chapters, and include quan titative comparis ons with the traditional field methods techniques. 4.2 Paleontology Paleontology and archeology field technique s require extensive digging and constant measurements. Usually at a di g site, before digging operation s begin, a regular grid is established. The positions of the fossils or arti facts that are discovered are carefully measured and recorded by photography. The current field me thods are very time consuming can really slow the dig progress specially when a very important finds such as articulate skeletons or unique artifacts are discovered. TLS systems can be used in these type of cases to provide both photographic and 3D measurements. Figure 4.1 illustr ate both the traditiona l and the alternative approach to obtain measurements at dig sites. Part A shows the traditional way were stakes and cordas are used to define the grid, while pa rt B shows a TLS measuring and recording a dig site.Digital terrain models from TLS data can provide additional information for monitoring progress such as dig volum e and cleared surface.

PAGE 63

63 Figure 4-1. Traditional and alte rnative methods for measuring a nd recording spatial information in archeological and paleontological site s. A) Traditional grid. B) TLS mapping. The paleontological application wa s tested at a Florida Museum of Natural History dig site located near Haile, Newberry; it is identified by Florida Museum as H7G. Figure 4-1b shows the ILRIS mounted on a tripod at that dig site. The site is a large sinkhole in a limestone quarry, containing fossil bones and teeth of about 40 diffe rent kinds of freshw ater and land animals dated at about 2 million years old (Pliocene E poch). Some fossils are preserved as intact skeletons but mainly isolated bones or teeth constitute the major finds. The study objectives were: 1) to establish a macro Geo-referenced 3D model of the dig site from several large scale ILRIS scans; 2) to obtain high density point clouds of small patches of the dig showing the fossils matched to the macro model and 3) if an interesting specimen was discovered the development of a multi angle high density point cloud model would be developed and embedded into the macro model. Figure 4-2 shows a point cloud rendering of th e macro model of the dig site. Figure 4-3 shows RGB textured renderings of high resolutio n point clouds obtained from a small patch of the dig site and Figure 4-4 presents before a nd after surface models generated from the point clouds. These were later used to determine the dig volume (~ 2.0 m3) after one day of excavation.

PAGE 64

64 Figure 4-2. Rendering of the dig site point cloud macro model s howing laser return intensity. Figure 4-3. RGB textured renderings of high resolution point clouds.

PAGE 65

65 Figure 4-4. 3D surface grids used to compute the volume of dirt extracted in one day. 4.3 Structural Geology Geology is a discipline that relies extensiv ely on mapping for the spatial recording of topography, crustal elements like faults and folds, stratigraph y and many other features. Plane table mapping is a traditional geological mapping technique, on which si mple tools (a flat leveled table, an angular measur ing device and a scale) are used to create large scale (1:120) maps. Plane table was replaced by modern survey ing equipment such as total stations and computer aided design (CAD) software. However, creating maps using the plane table or a total station are highly time consuming processes. Figure 4-5 shows the traditional table mapping technique and the alternative terr estrial laser scanning technique.

PAGE 66

66 Figure 4-5. Traditional and altern ative ways to perform geological field mapping. A) Traditional plane table mapping technique. B) Mode rn alternative mapping using TLS. TLS are ideal for creating geological maps for the study of geomorphology, stratigraphy and structural geomechanics as described by Un iversity of Texas Bureau of Economic Geology (Bellian et al., 2002). The tested application of TLS to structur al Geology was in support of NCALM PI Stephen Martel of the Department of Geology and Geophysics, University of Hawai’i at Manoa. Dr Martel has been developing theore tical models on the effect of topographic curvature on near-surfa ce stresses and the cr eation of sheeting join ts. Sheeting joints are opening mode rock fractures that form s ubparallel to the topographi c surface, develop to depths of at least 100 m and they occur main ly on regions where the topography is convex (Martel, 2006). The geology of Yosemite National Park in California exhibits vast areas with exposed sheeting joints; TLS mapping was perfor med on the Tuolumne quarry located along the Tioga road. The objectives of the project were 1) to map the sheeting joints on the exposed wall of the quarry on a 3D geodetic space and, 2) to the sheeting joints orientation to the topographic curvature of the area obtained from ALSM and th e mechanical stress to which the formation is subjected. Data was collected on September 22 and 23, 2006. The dataset consisted of 7 overlapping scans and vector observations for 6 reference G PS base stations that were used for the georeferencing of the data sets. The binary scan files were parsed to generate the XYZ & Laser

PAGE 67

67 Intensity and XYZ & RGB ASCII text files. Using Innovmetric Polyworks IMInspect module common points were identified in the overlapping scans to perform the merging operations. The coordinates of common points were input to Terrasolid Terras can to compute the solid 3D translation and rotation transformation paramete rs. After the transformation parameters were computed, the six sets of point clouds were tran sformed to the central scan coordinate system and a single point cloud was generated. Using Polyworks IMInspect module, the GPS cont rol points were iden tified in the merged point cloud and their sensor XYZ coordinates were determined. The GPS observation files were processed using the NGS Online Po sitioning User Service (OPUS) (http://www.ngs.noaa.gov/OPUS/) and the G PS coordinates for the control points were determined. The standard deviation of the coor dinate components varied between 0.014 to 0.789 meters, with a mean of 0.219 meters. With the sensor and UTM coordinates of the GPS control points; Terrascan was used to compute the so lid 3D translation and rotation transformation parameters for the data set. The RMS of the residuals of the transformation on the control points were 0.253919157, 0.156471707, 0.423406924 meters for th e Easting, Northing and Elevation components. The merged data set was geo-referenced and transformed to a UTM zone 11 (NAD_83) point cloud. Using Applied Imagery QT Modeler, intensity and RGB textured images were generated from the point cloud. In Figure 4-6 renderings of the point clouds from top and front view are presented. Using Terrascan, the point cloud was broken into tiles that later were imported to Golden Software Surfer to produce regular 2cm grids of the vertical walls. The triangulation with linear interpolation algorithm was selected for the grid creation. From the grids, shaded relief images were produced; Figure 4-7 shows one of su ch images where the layered sheeting joints

PAGE 68

68 are clearly seen. The images rendered from th e grids allows to determine the 3D spatial orientation of the joints with respec t to the surrounding topographic landscape. Figure 4-6. Geo-referenced poi nt cloud rendering of the Tuol umne quarry. A) Top view. B) Front view. Figure 4-7. Shaded relief image from a gri dded model of the sout h wall of the quarry.

PAGE 69

69 4.4 Wildlife Management Conservation Typical wildlife management and conservation activities include the catch and release of specimens for measuring, weighing and biol ogical sample collection. Under certain circumstances where there is a high risk of injury to wildlife personnel, when it is desired that the specimen not undergo the stress of capture or when just spatial measurements are needed a TLS may be an efficient alternative to the capture, me asure and release method. In this experiment an alligator was scanned from a safe distance of tw enty meters without pe rturbing the specimen. The results prove that is feasible to obtain accura te measurements of several dimensions such as length and thorax diameter. Volumetric models ca n also be created from the data, which provide a wealth of information for time series anal ysis. Figure 4-8 shows the traditional way of measuring alligators or crocodiles, it also s hows the rendering of a point cloud obtained by TLS of an alligator. From the point cl oud the alligator length was measured. Figure 4-8. Methods of measuring alligators and crocodiles. A) Traditional method. B) Alternative method from TLS point clouds. 4.5 Coastal Morphology The state of Florida has over 32000 kilometers of tidal shoreline of which more that 960 kilometers are beaches. The activities that are generated around these beaches, such as tourism, constitute a great source of income for the state. A large portion of the low-lying sandy beaches and dunes along the Atlantic coast are su bject to modification by high surf generated by

PAGE 70

70 northeasterly winds or by catastr ophic phenomena as tropical stor ms and hurricanes. Traditional methods of data collection and map generati on for beach profile change studies include differential leveling, traversi ng, static and kinematic GPS and aerial photogrammetry. These techniques are not only costly, time-consuming, and labor-intensive but also have poor spatial resolution (Shrestha, et al. 2005). Current Flor ida Department of Environmental Protection (FDEP) standards for beach profile topographic surveying require cr oss shore transects at Bureau of Beaches and Coastal Systems (BBCS) refere nce points, which are approximately 1,000 feet apart along shore, with a collection interval not to exceed 25 feet. and at all grade breaks and attributed items along the profile sufficient to accurately describe the topography at the profile locations (BBCC, 2004). These techniques do not have the spatial or temporal resolution required to precisely quantify and study the pro cesses of beach erosion, especially on erosion hotspots. Part A OF Figure 4-9 illustrate the us e of RTK GPS to generate beach transects and part B shows the UF M-TLS system us ed to generate beach surface maps. Figure 4-9. Methods for generati ng beach profiles. A) Using Real Time Kinematic (RTK) GPS. B) Alternative method using the M-TLS. The M-TLS was used to monitor an erosion hot spot located near the St. Augustine pier, Fl. at high spatial (cm level) and temporal (biweekl y) resolution. The methodology employed for data collection and analysis as well as the results are presented in Chapter 5.

PAGE 71

71 4.6 Soil Science In soil science; active and passive microwave re mote sensing techniques are applied to the derivation of soil parameters such as temperature and moisture. In order to derive soil moisture from radar sensors it is necessary to have a priori knowledge of the soil surface roughness. Traditionally soil roughness has been characterized as a single scale process obtained from 2D profiles and parameterized by the root mean square (RMS) of the height ( s ), the correlation length (l) and autocorrelation function (l( h )). The traditional methods of obtaining the soil pr ofiles are the Needle -like Profiler and the Mesh Board. A complete description of these mechanical profilers a nd the data collection procedures is provided by Mattia et al. (2003) The main disadvantage of these mechanical methods is that they tend to disturb the surface that is under study. Mesh Boards have to be hammered into the soil, while Needle-like Profil ers tend to penetrate into the surface yielding noisy measurements of the heights. Modern met hods are aimed to not disturb the surface; they are non-contact instruments. Th ese instruments include laser profilers, optical imagers and acoustic backscatter instruments (Mattia et al ., 2003; Zribi et al., 2000; Oelze et al., 2003). Figure 4.10 illustrate the meshboard and altern ate method for deriving soil roughness metrics. Figure 4-10. Methods for derivi ng soil roughness metrics. A) Tr aditional way using the mesh board. B) Alternative method using TLS to create 3D maps of the soil surface.

PAGE 72

72 The M-TLS is an excellent tool for soil roughness measurements because it has the capability of producing terrain models at sub-cen timeter scales with the additional advantage over traditional instruments and methods that it provides a complete surf ace digitizing, rather than digitized line profiles. The M-TLS was used to collect data in an experimental plot at the University of Florida Plant Science Research a nd Education Center at Ci tra, Florida and at a commercial plantation plot near Hastings, Fl. A detailed descri ption of the data reduction, analysis and result is presented in Chapter 6. 4.7 Forestry The estimation of forest structure and volume ha s for a long time been of great interest to the scientific community because of its ec ological and economic importance. Traditional methods of performing these measurements and es timations even over a small plot of forest require a great amount of man power and many h ours of field work. These methods require the manual measurement of tree height, stem diameter at breast height (dbh), stem location and stem density. Figure 4-11 part A shows the traditional way of measuring dbh, part B illustrate the use of a Biltmore stick to estimate tree height, part C show the alternative method of using TLS to generate point clouds from which dbh and tree height can be derived. Figure 4-11. Methods for estimating forestry me trics. A) Traditional method for measuring dbh. B) Traditional method for estimating tree he ight. C) Alternative method using the MTLS to create 3D maps of the forest.

PAGE 73

73 Since the early 1990s Airborne Laser Mapping (A LM) has been used to determine forest metrics. ALM provides a large spa tial coverage with very detailed 3D information of the forest upper canopy; however it provides very limited in formation of the forest understory structure and mass. Some experiments have proven that TLS can be used to provide high detail information on the understory with a limited spatia l coverage due to the line-of-sight obstruction caused by the same trees. An opportunity was iden tified for which dataset from both airborne and terrestrial platforms could be merged. Th e tested approach cons isted of developing techniques for geo-referencing TLS data set to achieve a seamless fusion with ALM data to generate high density point cloud of forest plots fr om which forestry metrics can be derived. This application is fully described in Chapter 7.

PAGE 74

74 CHAPTER 5 ST. AUGUSTINE BEACH EROSION HOT SPOT MAPPING 5.1 Motivation Costal engineers and scientists have known that beaches are subject to both natural or artificially induced sediment transport. With th e execution of large beach fills projects along the coasts of America in the 1980’s and early 1990’s, valuable experience was gained in long-term maintenance and beach-monitoring programs. That experience led to the recognition, systematic monitoring and study of Erosion Hot Spots or EHS. An EHS is an area that erodes more rapidly than the adjacent beaches or more rapidly than anticipated during beach fill design. Today knowledge of coastal processes is capable of expl aining what causes most types of EHSs and to formulate appropriate correction actions. EHSs can be classified and defined by several metrics such as loss of beach width (recession rate), lo ss of sediment volume (erosion rate), percentage of fill remaining of the amount placed, and perc eption of how a fill should perform relative to adjacent beaches or to historic rate (Kraus & Galgano, 2001). Airborne Laser Mapping technology has been ex tensively used to study large scale beach erosion. ALSM data covers a long stretch of beach with a moderate sample density of approximately 1 laser return per square meter (however, most current ALSM systems such as the Optech Gemini are capable of high pulse rates >100 kHz, with these type of systems 8 to 10 laser returns per square meter can be achieved). This sampling capability enables the detection of submeter-scale changes in shoreline position an d dune heights over peri ods of a few months. However, it might not be as effective for mappi ng short term, small scale variations that are characteristic of some localized erosion hot s pots. The M-TLSS, on the other hand, can provide high density point clouds (centime ter scale point spacing) of sm aller areas known to be highly

PAGE 75

75 prone to erosion. This chapter will discuss the application of M-TLSS as a complement to ALSM in the study of beach morphology in the St. Augustine, Florida area. 5.2 Use of LIDAR Technology Airborne LIDAR has been used since 1996 to study beach erosion. Early projects included the Airborne LIDAR Assessment of Coastal Eros ion (ALACE) that was a partnership between NOAA, NASA, and the USGS and the Laser Sw ath-mapping Evaluation and Resurvey (LASER) undertaken by the University of Florida, the Florida Department of Environmental Protection and the Florida Department of Transportation. Numerous papers have been published on the subject proving the success of this application. At the time of this writing, a literature search for the application of terrestrial laser scanners to study erosion hot spots yielded no results. The only reference wa s to a poster presentation at the 2005 meeting of The Geological Society of Amer ica (GSA). The abstract describes the use of a Terrestrial Laser Scanner to map a beach re-nou rishment plan covering 8.59 km of shoreline at Folly Beach, South Carolina (Kaufman et al. 2005). 5.3 Data Collection A known erosion hot spot (EHS) along the St. Augustine Beach, Fl. area was selected for this study. The EHS is located on the beach in fr ont of the St. Augustine Beach Front Resort (300 A1a Beach Blvd, St. Augustine, Fl.). Figure 5-1 s hows a map and a near infrared aerial photo of the study site, the orange polygon defines the mappe d area. Data were collected on 4 dates, one prior the beginning of the Hurricane season on Ma y 23, and 3 takes at two week intervals on October 28, November 10 and November 25, 2006. G PS data for the geo-referencing of the data set were also collected on the first three take s using Astech Z-Extreme and Astech Z-Surveyor geodetic grade receivers.

PAGE 76

76 Figure 5-1. Beach erosion hot spot study site location. A) Florida map, orange circle marks the location of St. Augustine. B) Aerial infr ared photograph of the study site, orange polygon marks the specific mapped area. 5.4 Data Processing The first step in the preliminary processing was the merging of individual scans taken for each day into a single point cloud in the sensor based coordinate system (XYZ). Polywork’s Inspect N-pair common point method was used fo r merging the scans. The next step was the geo-referencing of the point clouds; on the first take 7 GPS control stations were deployed, for these stations both XYZ and Easting, Northing, and Height coordinate s are available which allows for the computation of a 3D solid rota tion and translation transformation. Figure 5-2 shows a rendering of the March 23rd geo-refere nced point cloud, NAD83 was used as horizontal datum and NAVD88 as vertical datu m. In Table 5-1 the control po ints coordinates used for the geo-referencing transformation and the transfor mation residuals for the first data set are presented. To achieve comparable datasets the ge o-referencing of the last 3 point clouds were performed using 12 common points to the first ge o-referenced point cloud. The RMS from the process of coregistration of the point clouds based on the twelve common points are presented in Table 5-2.

PAGE 77

77 Figure 5-2. Rendering of the March 23rd, 2006 dataset. A set of control features that were used to visually verify the core gistration of the point clouds are presented in Figure 5-3. Different colors are used to identify point clouds collected on different days. From this figure it can be verifi ed the relatively good agre ement in coregistration among the scans as described by th e RMS values of Table 5-2. Table 5-1. Control points used for the geo -referencing of the Ma rch 23, 2006 dataset. Sensor Coordinates [m] Geodetic Coordinates [m] Transformation Residuals P X Y Z Easting Northing E Hgt Est Nrth Hgt P1 -27.497 -46.728 -0.623 474315.01 3303135.38 -23.59 -0.083 0.034 0.007 P2 -78.630 -32.978 -1.623 474346.58 3303177.83 -24.24 0.053 -0.002 -0.022 P3 5.096 63.037 -2.603 474404.80 3303064.44 -24.42 0.054 0.064 0.077 P4 58.391 108.561 1.111 474427.29 3302998.09 -20.61 0.088 0.067 0.366 P5 -166.835 -38.306 -1.419 474374.32 3303261.80 -23.89 -0.001 -0.066 0.039 P6 0.000 0.000 0.000 474348.44 3303092.52 -22.23 -0.284 0.050 -0.249 P7 32.780 171.860 0.959 474495.48 3302998.67 -19.34 0.174 -0.148 -0.218 RMS of Transformation Residuals [m] 0.137 0.074 0.190 Table 5-2. Point clouds coregistration RMS values. Coregistration RMS [m] Point Clouds Easting Northing Height 1-2 0.108 0.237 0.135 1-3 0.090 0.211 0.116 1-4 0.086 0.133 0.098

PAGE 78

78 Figure 5-3. Features used to check th e co registration of the point clouds. The third step consisted of cr opping the point clouds to the specific area of interest as defined by the orange polygon on Figure 5-1 B and filtering to remove all the non-surface objects (people, beach chairs, etc.). The fourth and final step was the creation of 10 cm spacing regular grids by the method of triangulation with linear interpol ation using Surfer & Matlab software packages. Figure 5-4 present image ma ps created from the 10 cm elevation grids. Figure 5-4. Image maps from the 10 cm elevation grids.

PAGE 79

79 5.5 Results 5.5.1 Elevation Changes Elevation changes were computed for both s hort and long term peri ods. Using Surfer the elevation grids were differenced two at a time Two comparisons were performed for long term change (May 23 to October 28 and May 23 to N ovember 25) and two for short term change (October 28 to November 10 and November 10 to November 25). The results of the elevation change detection are presented as difference gr ids on Figure 5-5. The long term change grids reflect an average of 20 cm of reduction in the berm elevation and a 1.4 to 2 m difference in elevation between the berm and the surf zone. The short term change grids show a general preservation of the berm elevation and an averag e difference of 1m in the berm to surf zone elevation. Figure 5-5. Image maps from th e elevation change grids.

PAGE 80

80 5.5.2 Volume Changes Total volume differences were computed and normalized to obtain lost volumes per unit beach length and rates of lost volume per unit be ach length. For planners and engineers the total volume lost is important because that translates directly to the renourishment costs, however absolute measurements are difficult to use fo r comparisons; lost volumes per unit beach length and rates of lost volume per unit beach length to provide a better understa nding of the magnitude of the change. Table 5-3 summarizes some of the volume computations performed from the elevation grids. Table 5-3. Summary of volume change computations. Lost volume Volume loss rate Volume loss/beach unit length Volume loss/beach unit length rate From To Days [m] [m/day] [m/m] [m/m day] 5/23/2006 11/25/2006 186 13168.245 70.79702 59.31642 0.318905 5/23/2006 10/28/2006 158 9166.72 58.01722 41.29153 0.261339 10/28/2006 11/25/2006 28 4945.707 176.6324 22.27796 0.795641 10/28/2006 11/10/2006 13 5535.53 425.81 24.93482 1.918063 5.5.3 Beach Line and Crest of Berm Extraction From the Grids The recession of the beach line is an impor tant phenomenon to record and quantify. The beach line is defined by the mean higher high wa ter (MHHW) line; which is an average of the higher high water height of each tidal day over nearly 19 years. For the study area the mean higher high water line is determined at 0.6 me ters for the NAVD88 vertical datum. Because the elevation grids were created usi ng ellipsoids heights a conversi on from orthometric height to ellipsoidal height was performed. The average Geoid separation was found to be 28.612 meters so the beach line was extracted from the -28.012 me ters ellipsoidal height contour. Similarly the crest of berm was extracted from the -26.5 mete rs contour, Figure 5-6 co ntains plots with the beach and crest of berm lines.

PAGE 81

81 Figure 5-6. Beach line and crest of berm position plots for each of the data collection dates. 5.5.4 Across Beach Profile Extraction The traditional data collection method for studying beach erosion is transect sampling. FDEP standard require cross s hore transects at every 1,000 feet apart along shore; with a collection interval not to exceed 25 feet and at all grade br eaks. From the M-TLS dataset generated grids, transects can be obtained automa tically at higher resolutions such as presented in Figure 5-7. To do any kind of interpretation w ith this transects is necessary to specify a tolerance in the horizontal and vertical components due to the accuracy of the coregistration procedure. This tolerance is in the same order of the highest residual of the coregistration control points. These residuals were in the order of 11cm in the East-Wes t direction, 24 cm in the NorthSouth direction and 13 cm in the vertical dimensi on. From these transects it can be seen that the beach line and the berm line reced ed twenty meters on average. However, the most interesting

PAGE 82

82 result is that half of that r ecession occurred in the last four we ek period of the complete 27 week observation program. This accelerat ed erosion can be related to the appearance of the North Eastern winds which are recognized as the main se diment transport mechanism of that area. The berm maintained a constant height of roughly 3 meters above the beach line. Figure 5-7. Beach profiles ex tracted from the grids showi ng the recession of the berm.

PAGE 83

83 5.6 Comparison Between Traditional Methods and M-TLS Based on the resolution established by FDEP standards for beach profile topographic surveying, simulated GPS or leveling profiles were generated to illustrate the difference in resolution between traditional surveying methods and these achieved by the M-TLS. The results are presented in Figure 5-8, th ese show that the tr aditional methods do no t capture the small scale details of the beach and be rm surface. It can also be seen that the traditional method can over estimate beach erosion as they do not sample properly the berm crest. Figure 5-8. Comparison of profile resolution generated from traditional methods and M-TLS.

PAGE 84

84 CHAPTER 6 SOIL ROUGHNESS METRICS DETERMINATION 6.1 Motivation The application of active and pa ssive microwave remote sensi ng is increasing in the field of soil science as a tool to map soil properties. It is possible to use mi crowave backscattering to extract geophysical surface parameters such as soil moisture content and soil surface roughness. The scattering of microwave energy is determined by the sensor parameters such as wavelength, polarization and observation geometry; it also depends on the surface dielectric properties and roughness. In theory it is a simple computation to solve for one of the parameters (moisture or roughness) having prior knowledge of the other a nd of the microwave energy backscattering. However many practical problem arise when tr ying to parameterize the soil surface roughness. (Callens et al.; 2006 & Zribi et al.; 2000) Trad itionally soil roughness has been ch aracterized as a single scale process obtained fr om 2D profiles and parameteri zed by the root mean square (RMS) height ( s ), the correlation length (l) and autocorrelation function ( p (l)). The main limitations of this modeling as reported in the literature are: The soil surface and its roughness are multiscale in nature, simplifying them to single scale parameters implies a loss of information. Theoretical and field data have shown that different values of the roughness parameters, especially on “ s ” and “l”, can be obtained from the same surface as a function of the profile length, discretization interval, the instrume nt resolution, and the overall shape of the profile. The surfaces are assumed to follow Exponential or Gaussian distribut ions without having the ability to check these assumptions. To overcome these limitations and the inadequacy of the single scale models in describing complex soil surfaces, several alternative multis cale roughness description models have been proposed. They include the mixture of small and large single scale features, the use of random fractals and fractal dimensions to describe th e surface (Davidson et al., 2000), and the use of

PAGE 85

85 plane facets and 3D statistical analysis (Zribi et al., 2000). Ho wever, the universally accepted theoretical microwave backsca ttering models such as the Small Perturbation Model (SPM), Kirchoff Approximation (KA) and th e Integral Equation Model (IEM) continue to require only the single scale parameters as inputs. A novel approach using the M-TLS to scan sc attering surfaces to generate 3D terrain models was explored. From the 3D models the distribution function of the single scale parameters can be obtained, which is expected to fully describe th e surface roughness, thus overcoming the limitation of under representati on produced by the profiling sampling methods. 6.2 Use of LIDAR Technology The traditional methods of obtaining the soil pr ofiles to compute the roughness parameters by mechanical means are the Need le-like Profiler and the Mesh Board. A complete description of these mechanical profilers and the data colle ction procedures is provided by Mattia et al. (2003). The main disadvantage of these mechanical methods is that they tend to disturb the surface that is under study. Mesh Boards have to be hammered into the soil, while Needle-like Profilers tend to penetrate into the surface yielding noisy measurements of the heights. Modern methods are aimed at not disturbing the surface; these methods use non-contact instruments such as laser profilers optical imagers and acoustic b ackscatter instruments (Mat tia et al., 2003; Zribi et al., 2000; Oelze et al., 2003). Laser profilers have been used intensely fo r soil roughness digitizing; most of them are commercial systems capable of high spatial resolu tions of the order of 1 mm or less but limited to relatively short profile lengths, usually no more than a few meters. Special laser profilers have been develop to allow the measurement of l onger profiles, such as the CESBIO-ESA laser profiler, which is capable of acquiring r oughness profiles up to 25 m long (Davidson et al., 2000).

PAGE 86

86 Whether using mechanical or non-contact techniques, there ar e sources of error that are common when characterizing the soil roughness. These are the tr uncation Error, which arises from measuring relatively short profiles; and profiler error, which is due to the intrinsic limitations of a measurement method (Mattia et al., 2003). It is expected that by full surface digitizing, using the M-TLS, the errors from these sources will be drastically reduced. A search of the literature found no journal papers on the use of terrestrial laser scanne r to digitize a surface and later compute its roughness parameters. 6.3 Data Collection Data was collected using the ILRIS in an expe rimental plot at the University of Florida Plant Science Research and Educa tion Center at Citra, Florida. The collected data contains the soil surface of the footprints of two passive microwave radiom eters operating in the L and C bands. Three different time samples were collected, th e first just after soil tilling, the second after corn planting and the third after the crop was harv ested. An additional data set that contains a larger horizontal variation was collected on a co mmercial plantation near Hastings, Florida. A description of the collected da ta is provided in Table 6-1. Table 6-1. Soil roughness collected datasets Dataset ID Collection Date C onditions Radiometer Footprint Field1 Field3 March 08, 2006 After Tilling Mesh Board Test Citra01 L Citra02 C Citra03 Mesh Board Test Citra04 Mesh Board Test Citra05 L Citra06 March 10, 2006 After Planting C CitraCBand C CitraLBand May 30, 2006 After Harvesting L Hastings October 28, 2006 Plowed and Planted N/A

PAGE 87

87 Two traditional meshboard measurements were made near the L Band and C Band radiometer footprints with the soil tilled. For the present analysis only two data sets will be considered: Citra 02 and Hastings. 6.4 Data Processing The first step in the process was to prepare the point clouds for analysis. For each test area this consisted of converting the ex isting tilted point cloud terrain into that of a flat level terrain. The angles required to perform the transformation were determined by fitting a plane to selected ground points, finding the normal v ector and computing the rotati on angles about the X and Y axes. The rotation was performed using TerraSc an transformation module. Once leveled the point cloud was rotated about the Z axis to align the radiometer f ootprint axes to the point cloud X and Y axes. Finally the point cloud was cropped to a 4 x 6 me ter plot corresponding to the radiometer footprint. Figure 6-1 shows renderi ngs of the raw and a rectified and point cloud. Figure 6-1. Dataset preprocessing steps. Rendering of A) raw point clouds and B) rectified point cloud from the Citra 02.

PAGE 88

88 From the cropped and rectified point clouds regular grids with one cm cell spacing were created using a triangula tion with linear interpol ation gridding function in Matlab. Figure 6-2 shows image maps of the elevation grid s for two of the analyzed dataset. Figure 6-2. Renderings of th e 1 cm elevation grids. A) Citra 02. B) Hastings. The cropped point clouds and the grids were us ed to derive the roughness metrics. The expressions for the height RMS, correlation le ngth and autocorrelation function for 2D profiles are given in Table 6-2. Several tests were performed to prove the advantages of using MTLS datasets to extract the single sc ale metrics over the traditional pr ofiling methods. These tests are: From the regular grid a set of random 2D profiles was extracted parallel to the X and Y axes. The root mean square height (s) and the correl ation length (l) for each profile were computed using the traditional formulas. Mean and standard deviations of the roughness parameters of the profiles of the same plot were computed. The formulas used to compute root mean square height (s) and correlation length (l) from the 2D transects were extended to 3D surfaces and are used to compute the roughness metrics for

PAGE 89

89 the two plots. The correlation was also exte nded from 1D to 2D, so the correlation length could be found for either X and Y, or a comb ination of both directi ons. The 2D correlation length was converted into a 1D length and dist ribution functions of correlation lengths were obtained. Comparisons were made between the averag ed roughness parameters values from the random profiles with the ones obtained from the 3D models. Test were performed to verify the assumpti on that the autocorrelation functions follow Exponential or Gaussian forms. Table 6-2 Definition of soil roughness parameters. Callens et al. Thoma et al. Height mean N i iz N z11 Height RMS 2 1 21 1z N z N sN i i N i iz z N s1 21Normalized correlation function N i i j N i j i iz z z x j h1 2 1 N i i j N i j i iz z z z z z x j h1 2 1 Correlation length e l that such l 1 : Exponential autocorrelation function l he h Gaussian autocorrelation function 2 2l he h 6.5 Results 6.5.1 Simulated Profiling Results From the elevation grids random transects were extracted parallel to the X & Y axes to simulate traditional soil roughness profiling tech niques. Using the 2D formulas the roughness metrics were computed to verify the variability of the metrics with the profile selection. Figures 6-4 to 6-7 show the extracted profiles and thei r respective normalized au tocorrelation plot from the Citra 02 and Hasting datasets. The red line in the autocorrelation plots represents the 1/e value used to determine the correlation length.

PAGE 90

90 Figure 6-4. Roughness parameter plots for the C itra 02 dataset parallel to the X axis. A) Extracted profiles. B) Normalized autocorrelation plot. Figure 6-5. Roughness parameter plots for the C itra 02 dataset parallel to the Y axis. A) Extracted profiles. B) Normalized autocorrelation plot. Figure 6-6. Roughness parameter plots for the Has tings dataset parallel to the X axis. A) Extracted profiles. B) Normalized autocorrelation plot.

PAGE 91

91 Figure 6-7. Roughness parameter plots for the Has tings dataset parallel to the Y axis. A) Extracted profiles. B) Normalized autocorrelation plot. The results of the roughness metrics are summari zed in Table 6-3 for the Citra 02 data set and in Table 6-4 for Hastings. It can be observed from the tables that the mean height RMS from profiles in both directions (parallel to Y axis and parallel to X axis) are relatively close (3 to 25% variation), however the correlation length can vary greatly between directions (39-77%). Table 6-3. Soil roughness parameters results from random profiles fo r the Citra 02 grid. Data set RMS regular spacing [m] 1/e correlation length [m] Citra 02 Y=1 m 0.0133 2.1718 Citra 02 Y=3 m 0.0100 2.1657 Citra 02 Y=5 m 0.0108 2.3727 Citra 02 X=0.75 m 0.0115 4.0464 Citra 02 X=1.75 m 0.0079 3.8960 Citra 02 X=2.75 m 0.0077 3.9844 Mean parallel to X 0.0114 2.2367 Mean parallel to Y 0.0090 3.9756 Overall mean 0.0102 3.1062 Overall 0.0022 0.9565 Table 6-4. Soil roughness parameters results from random profiles for the Hastings grid. Data Set RMS Regular Spacing [m] 1/e Correlation Length [m] Hastings Y=1 m 0.0277 1.9087 Hastings Y=3 m 0.0543 2.0181 Hastings Y=5 m 0.0715 2.0127 Hastings X=1 m 0.0458 2.6581 Hastings X=3 m 0.0497 2.9350 Hastings X=5 m 0.0633 2.6739 Mean parallel to X 0.051 1.980 Mean parallel to Y 0.053 2.756 Overall Mean 0.052 2.368

PAGE 92

92 6.5.2 Extension of the 2D Formulas for a 3D Surface. The height RMS formula provided for 2D pr ofiles in Table 62 is valid without modification for the 3D soil surfac e. For 2D profiles the correlat ion length is defined as the distance for which the value of the normalized 1D autocorrelation is 1/e. The concept can be extended for a 3D surface by computing a 2D auto correlation with two parameters (x lag and y lag) which results in a 3D surface. The 1/e co ntour can be located on that surface, and the correlation length can be defined as the scaled di stance (considering grid element size) from the origin to a particular point on the contour. Th e advantage of this method over the traditional profiling method is that from this a complete distribution function of the correlation length is obtained rather than a single value. Figure 6-8. Citra normalized height autocorrelation. A) 3D plot. B) Color map. C) Contour plot.

PAGE 93

93 Figure 6-9. Hastings height au tocorrelation. A) 3D plot. B) Color map. C) Contour plot. Figure 6-10. Correlation lengths extr action for the Citra 02 dataset. A) 1/e contour plot of the 2D autocorrelation function. B) Unfolded corre lation length plot. C) Distribution function of correlation lengths.

PAGE 94

94 Figure 6-11. Correlation lengths extr action for the Hastings dataset. A) 1/e contour plot of the 2D autocorrelation function. B) Unfolded corre lation length plot. C) Distribution function of correlation lengths. Graphs of the 2D autocorrelation are presented in figures 6.8 and 6.9, part A of the figures are 3D plots that show the normalized autocorre lation function as a function of the X and Y lag units. Part B is an image map of the same norma lized autocorrelation values and part C is a contour plot of the correlation va lues. From the normalized auto correlation the 1/e contour can be extracted as shown in figur es 6.10 and 6.11 part A. From each point in the contour its direction and correlation length ca n be computed. In part B of figures 6.10 and 6.11 the contour is unfolded and the correlation length for each po int is computed and plo tted without taking into consideration the direction. The obtained co rrelation length are then grouped and binned to create the histograms on part C of the figures Table 6-5 provides a summary of the roughness parameters values extracted from the 3D surface models. From this table and figures 6.10C and

PAGE 95

95 6.11C it can be seen that the mean correlation leng th is not necessarily a good descriptor of the distribution. Table 6-5. Soil roughness parameters from 3D surface models Roughness Parameter Citra 02 Hastings (s) RMS height from point cloud [m] 0.013546 0.05013 (s) RMS height from grid model [m] 0.012674 0.0547 ( l ) Mean correlation length [grid units] 271.78 215.25 ( l ) Mean correlation length [m] 2.7178 2.1525 Standard deviation of correlati on length [grid units] 60.42 52.23 Standard deviation of corre lation length [m] 0.6042 0.5223 6.5.3 Comparisons of Roughness Metrics From Profiles vs. Full Surface The values of the roughness parameters obtained from the random profiles and the 3D surface are tabulated for comparison in tables 6.6 and 6.7. Table 6-6. Comparison of soil roughness paramete rs for Citra 02 from 3D surface models and random generated profiles. Roughness Parameter 3D Surface Transects // to X Transects // to Y All transects (s) RMS height [m] 0.012674 0.0114 0.0090 0.0102 ( l ) mean correlation length [m] 2.7178 2.2367 3.9756 3.1062 Min correlation length [m] 2.0597 Max correlation length [m] 3.9651 Mode correlation length [m] 2.2 of correlation length [m] 0.6042 0.1178 0.0756 0.9565 Table 6-7. Comparison of soil roughness paramete rs for Hastings 02 from 3D surface models and random generated profiles. Roughness Parameter 3D Surface Transects // to X Transects // to Y All transects (s) RMS height [m] 0.0547 0.051 0.053 0.052 ( l ) mean correlation length [m] 2.1525 1.980 2.756 2.368 Min correlation length [m] 1.4200 Max correlation length [m] 3.1620 Mode correlation length [m] 2.1 of correlation length [m] 0.5223 0.062 0.156 0.438 6.5.4 Distribution Functions of the 3D Correlations Lengths. From Table 6-2 it can be seen that the distributions of the correlation lengths are assumed to follow either and exponential autocorrelati on function (ACF) or a Gaussian autocorrelation function form. One of the advantages of the full surface digitizing is that the complete

PAGE 96

96 distribution function of the co rrelation lengths can be obt ained. Figure 6-12 shows the distribution and compares it with a normal Gaussian distribution. It can be clearly seen that the distribution is not close to either an exponential or Gaussian distributions. Figure 6-12. Comparison of experi mental correlation le ngth distributions with respect to the assumed normal distribution. A) for the Citr a 02 dataset B) for the Hastings dataset. 6.6 Comparison with the Traditional Meshboard Method. On March 10, 2006 meshboard measurements were made near the footprints of the radiometers. Figure 6-13 illustrate the meshboard technique, a picture is taken of the interface between the soil and the board. Imag e processing software is used to digitize the coordinates of the soil surface based on the mesh printed on the board. Figure 6-13. Meshboard used to di gitize the soil surface transect.

PAGE 97

97 Figure 6-14 show the digitized soil surface tran sect from one of the footprint areas based on the meshboard method. On part A the irregular spaced points are plotted, part B plots a regular spacing (1 cm) sampling interpolation. Part C is a plot of the normalized 1D autocorrelation. Figure 6-14. Plots of meshboard derived data. A) irregular sampled prof ile B) regular spacing profile C) normalized height autocorrelation. The results obtained from the traditional mes hboard method are compared with the results obtained thru the 3D datasets in Table 6-8. Ther e is a disagreement in the values obtained from the different methods, being the greatest diffe rence the autocorrelation length on the C-band radiometer footprint. Table 6-8 Comparison of soil roughness metrics obtained from the trad itional and alternative method. Data Set RMS all points RMS regular spacing 1/e correlation length MB Near L Band FPT 0.00915 8 m 0.009291 m 1.427 m 3D L Band FTP 0.02179 m 0.022513 m 2.2512 m 2. Near C Band FPT 0.01311 7 m 0.012122 m 0.284 m 3D C Band FTP 0.013546 m 0.012674 m 2.7244 m

PAGE 98

98 6.7 Conclusions Having a 3D data set of a scattering surface allowed us to demonstrate the common sense knowledge that the values of the derived so il roughness parameters are highly variable depending on the profile selection. It was also possible to derive the traditional 2D profile metrics from an entire 3D surface, having not a single value for correlation length but a range of values that fully describe the scattering surf ace. This proved the advantage of a full surface modeling over the tradi tional profiling methods.

PAGE 99

99 CHAPTER 7 FORESTRY METRICS APPLICATIONS 7.1 Motivation The estimation of forest structure and volume ha s for decades been of great interest to the scientific community because of its ecological and economical importance. Traditional methods of performing these measurements and estimations even over a small plot of forest, are labor intensive and require many hours of field wo rk. Several non-invasive remote sensing technologies have been tried to simplify these ta sks and collect more accurate measurements in less time. Over the past several years a great amount of experience has been obtained in the application of Airborne Lase r Mapping (ALM) to study forest structure and estimating its biomass. ALM provides a large spa tial coverage with very detailed 3D information of the forest upper canopy; however, it provides very limited information of th e forest understory structure and mass. On the other hand, Ground-Based S canning Laser or Terres trial Laser Mapping (TLM) can provide very detailed 3D informati on on the understory struct ure with limited spatial coverage due to the line-of-sight obstruction caused by the trees. In contrast to ALM, which is a relatively mature technology with commonl y accepted data collecting techniques and procedures, there are no guidelines or uniform ly accepted procedures on how to set up a TLM system to obtain data for forestry applications. Bibliographical research indica tes that very few efforts have been made to use TLM to investigate Forest Metric s and structure. Previous work has b een limited to assess the potential of TLM for replacing traditional field techniques to determine basic tree & plot metrics such as tree height, stem diameter at breas t height (DBH), stem location and stem density; or as ground validation for airborne remo te sensing technologies.

PAGE 100

100 The long term goal of this proj ect is to develop techniques for a seamless merging of ALM and TLS observations to generate high density poin t clouds of forest plots. The final goal is to develop high resolution laser tomography throug h a multi-platform and multi-imaging-geometry data integration to generate a virtual “Forest Cube” for which all metrics can be derived with high levels of accuracy and precision. With the derived metrics of several of these “Forrest Cubes” of statistically sampled plots combined w ith the large spatial coverage of ALM, forestwide metrics can be estimated with a high degree of confidence. The first step and shorter term goal is to develop the required fi eld techniques and procedures fo r TLM forestry data acquisition and geo-referencing for a typi cal small size test plot. 7.2 Use of LIDAR Technology In the early 1980s LIDAR scanners we re developed to obtain high accurate measurements of surface elevations from airbor ne and spaceborne platforms. These first systems are classified as large footprint LIDARs because the intersection of the laser beam with the surface is a circular or elliptica l spot of several meters in extent. The first LIDAR systems suitable for vegetation and forestry studies had a footprint of less than a meter and were designed to record only the first received return (Lefsky et al. 1998). Those systems evolved to present day airborne laser scanners that generally have small footprint and the capability of recording multiple returns, as well as their intensities Detailed descriptions of LIDAR technologies instrumentation and operation can be found in Wehr & Lohr (1999) and Baltsavias (1999). Ground-Based Laser Scanning or Terrestrial La ser Scanning (TLS), is a relatively new technology compared with their fl ying counterparts. They evolved from the Electronic Distance Measurement (EDM) devices used in traditional surveying. Terrestrial laser mapping systems produce very accurate 3D data sets of la rge surfaces rapidly. Other surveying or photogrammetric techniques require much longe r acquisition times and yield much fewer 3D

PAGE 101

101 points. A ground-based laser sca nner provides 4 quantities for each sampled point, three are positional information in a Cartesian reference fr ame: the (x,y,z) coordinates, and the fourth quantity is the magnitude or intensity of the retu rn signal. The sampling resolution is measured in angular units because linear spacing between poi nts depends on the range. For a 2 milliradian angular resolution the linear spac ing between samples is 1 cm at 10m or 10 cm at 100 meters. The key features of a ground-based laser scanni ng system are: a) ra nge accuracy: which is dependent on the pulse durati on, b) maximum range: which de pends on the laser output power and the receiver sensitivity, c) scan rate: de pends on how fast are the laser and receiving electronics, d) the angular resolu tion, e) field of view: the last two depend on the mechanics of the scanning mirror system, and the last but most im portant is the f) laser, which is characterized by its wavelength and beam divergence. More detailed information on TLS can be found in Lichti et al. (2002). 7.3 Previous Works Over the last two decades multiple papers on Airborne LIDAR applications to vegetation and forest studies have been publis hed, discussing the capabilities and limitations of the different systems and applications. Some of the most comple te and interesting papers are: Lefsky, et al. (1998); Lefsky et al. (2002) a nd Nsset & Gobakken (2005). The overall conclusion is that airborne LIDAR mapping is a ve ry powerful tool for deriving and modeling forest metrics; however, it has the limitation of the lo w sampling of the understory canopy. Previous works on the application of TLS to forestry have aimed to compare tree & plot metrics such as tree height, stem diameter at br east height (dbh), stem location and stem density derived from TLS with the same metrics deri ved with traditional ground techniques. These works have had different approaches with resp ect to scanning geometry and the shape of the defined plot. Watt & Donoghue define d circular plots of 0.02 hect are (ha) and obtained a quasi-

PAGE 102

102 complete scan of the area with only two scans from opposite points. Hopk inson et al. defined 35 m x 35m (0.12 ha) square plots a nd performed 5 scans per plot. Bo th studies concluded that TLS was a powerful tool in determining forest variab les; however it is limite d by obstructions of the line of sight that determines the useful range of this type of instrument. The above papers prove the usefulness of bot h airborne and terrestri al based systems in the estimation of forest metrics. The work presen ted here is aimed at de veloping data acquisition and processing techniques that will enable the combination of data from multiple platforms (i.e. Airborne & Terrestrial) and geom etries that will produce a complete well sampled data set over the complete structure of a small plot of the forest. 7.4 Data Collection The test site for the experiment was at the Intensive Management Practice Assessment Center (IMPAC) operated by the Forest Biology Research Cooperative (FBRC), and located 10 km north of Gainesville, Florida USA. The center is a research plantation of the southern pine species loblolly (Pinus taeda L.) and slash (Pinus elliottii var. elliottii). The Airborne Laser data was collected on October, 2005 using the UF Optech Inc. ALTM 1233 laser mapping system flown on a Cessna 337. The airborne system, when ope rating at a flight hei ght of 600 m AGL, at a flight speed of 60 m/s and a LA SER pulse repetition frequency of 33 kHz yields a point density of roughly 1 laser return per square meter. Two sets of terrestria l Laser Data were collected, the first in November 2005 and the second in June of 2006 using the UF Mobile Terrestrial Laser Scanner (M-TLS) system based on the Optech IL RIS-3D terrestrial im aging LIDAR. Figure 7-1 presents an aerial photo of the te st site, and Figure 7-2 presents a shaded relief model rendered from the airborne laser scanner data.

PAGE 103

103 Figure 7-1. Aerial photographs of the test site. Downloaded from the Florida Department of Environmental Protection, Land Boundary Information System. http://data.labins.org/ 2003/MappingData/DOQQ/doqq.cfm The terrestrial laser mapping data consisted of five independent scans taken from different angles and elevations. The tota l number of return poi nts in the terrestrial clouds was 3,765,084. The TLM instrument provides a set of points (x,y ,z,I). The first three poi nts provide the spatial information of the scanned surface point. These x,y,z coordinates are distances referred on an orthogonal frame of reference whose origin is the scanner sensor head. Geo-referencing was accomplished by setting GPS control points that were included in the scans, and later determining the geodetic coordi nates of these control points. Data was collected using ASHTECH Z-XTREME receivers, and ASHTEC H Choke Ring Antennas Model 700936 Rev D. A total of 8 stations were surveyed, 6 corres ponding to the GPS contro l points and 2 for the scanning locations.

PAGE 104

104 Figure 7-2. Shaded relief digital elevation model rendered from the airborne laser scanner data of the test site. The edge eff ects are caused by the method used to interpolate the LIDAR observed data to obtain a uniform spaced (gridded) surface model. 7.5 Data Processing The five individual point clouds were merged into a single data set in a single sensor coordinate frame using the Innovmetric (http://www.innovmetric.com/Manuf acturing/home.aspx) Polywork s Inspect software employing the n common point alignment procedure. The G PS reference point (ARP) were used as the reference points for the alignment. For georeferencing the dataset into a NAD83 Datum expressed in UTM Zone 17 coordinates, 8 contro l points were surveyed. Their coordinates in both geodetic and sensor spaces ar e listed in the Table 7-1. Table 7-1. M-TLS data set georeferencing control network. Geodetic space coordinates UTM zone 17 Sensor space coordinates Station Easting (X) Northing (Y) Ellipsoid Height (Z) x y z GPS1 374997.280 3293231.190 26.067 1.156 21.033 -2.505 GPS2 375004.345 3293235.345 24.982 -6.77 19.81 -4.403 GPS3 375014.622 3293203.487 35.906 -4.275 54.339 -2.766 GPS4 374994.418 3293223.649 25.736 7.06 26.576 -3.955 GPS5 375011.541 3293247.681 25.621 -18.4871 11.63806 -3.71564 GPS6 375011.129 3293242.494 23.451 -15.7089 16.51747 -5.505 IL01 374989.820 3293251.056 23.701 0 0 0 IL0_ 375015.927 3293268.460 24.739 -30.9398 -5.44228 -2.60267

PAGE 105

105 Following, a 3D translate and rotate conf ormal transformation was applied using the Terrasolid (http://www.terrasolid.fi/ ) Terrascan software. The inputs for the computation of the 9 parameters were the sensor space and geodetic space coordinates of the GPS antennas reference point (ARP). In Table 7-2 the residuals of th e control points based on the geo-referencing transformation are presented. Table 7-2. Geo-referencing residuals analysis. Adjustment Residuals Square of the Residuals Station Easting (X) Northing (Y) Ellipsoid Height (Z) x y z GPS1 -0.0736 0.0591 0.3474 0.005417 0.003493 0.120687 GPS2 -0.0198 0.0011 0.4919 0.000392 1.21E-06 0.241966 GPS3 0.0459 0.2654 -0.3605 0.002107 0.070437 0.12996 GPS4 -0.0541 -0.0475 -0.2618 0.002927 0.002256 0.068539 GPS5 0.0493 -0.0022 0.1346 0.00243 4.84E-06 0.018117 GPS6 This control point was eliminated after a preliminary adjustment IL01 -0.055 -0.1356 -0.1181 0.003025 0.018387 0.013948 IL0_ 0.1073 -0.1404 -0.2336 0.011513 0.019712 0.054569 Root Mean Square of the Residuals (Meters) 0.063032 0.127779 0.304205 Once the Terrestrial Point cloud was geo-refere nced in the same datum and coordinate system as the airborne data, both datasets were merged into a single file and were viewed and rendered. Figure 7-3 is a rendering of the fused poi nt cloud color coded by elevation; the yellow ellipse indicates the overlaying of the high density terrestrial point cloud. It can be observed the adequate mapping of the undercanopy structure by the high density of the blue and green color coded points. Figure 7-4 is a rend ering of a 3.0 m wide transect depicting the match between the airborne and the terrestrial point clouds in the along flightline directi on, red dots represent the low density points obtained from the airborne scanner and the white dot s represents the high density terrestrial point cloud. Figure 7-5 illustrate the match between the airborne and the terrestrial point clouds in the across flightline direction.

PAGE 106

106 Figure 7-3. Rendering of the fused point cloud, color coded by elevation. Figure 7-4. Rendering of fused poi nt cloud cross section in the along the flightline direction. Figure 7-5. Rendering of fused point cloud cross section in th e cross flightline direction

PAGE 107

107 Figures 7.6 to 7.8 are renderings of the fuse d airborne and M-TLS point cloud, some of the GPS tripods and Antennas that were used as c ontrol points can be obser ved in the renderings. Figure 7-6. Rendering of the fuse d point cloud, grey scale from the laser return intensity. Figure 7-7. Rendering of the fused point cloud, color coded by elev ation + laser re turn intensity.

PAGE 108

108 Figure 7-8. Rendering of the top view of fused point cloud, colo r coded by elevation + laser return intensity 7.6 Results The objective of merging the ai rborne and terrestrial point clouds was to synergize the vantage points of both geometries to develop a dataset that could permit the extraction of the Tree & Plot Metrics (Tree Height, Stem Diameter at Breast Height DBH, Stem location and Stem Density) with higher levels of precision and accuracy than those using a single geometry. From the fused point cloud a “Forest Cube” wa s extracted, its ground dimensions are 40.4m x 19.16m with an area of 774.15 m. From this virtual “Cube” all the metrics will be derived. Figure 7-9 shows renderings of the “Forest Cube”. 7.6.1 Stem Density For stem density determination, a manual coun t of the standing stems was performed. To facilitate the count the “cube” was skimmed to 1/20th of its point density, was filtered so that only points that had heights between 2.3 and 6. 3 m above ground level were displayed and was visualized in an oblique view Figure 7-10 shows the stem coun t by row, a total of 96 standing stems were counted, which yield a stem density of 1 stem per 8.06 m or 1240 stems/hectare.

PAGE 109

109 Figure 7-9. Rendering of the “Fores t Cube”. A) Color coded by elevation and intensity. B) Color coded only by elevation. Figure 7-10. Rendering of point cloud used for stem counts.

PAGE 110

110 7.6.2 Stem Location Previous published works have also performe d stem location measurements, however they were referred to arbitrary sensor coordinates. In this work, the dataset was referenced to geodetic coordinates, providing absolute c oordinates of the stem location. The ideal approach is to fit a circle to a low cross section of the tree and dete rmine the coordinates of the circle origin and the circle radius. Polyworks INspect or Terra Scan can be used for this purpose. 7.6.3 Stem Diameter at Breast Height DBH Stem diameter at breast height is defined as the stem diameter at a hei ght of 1.4 m (4.5 ft). The stem diameter can be determined in conjunction of the stem location. Figure 7-11 illustrates the fitting of a circle from a section of a point cloud at breast height, from which the stem locations and DBH were determined. Figure 7-11. Fitting of a circle at breast he ight for determining DBH and stem location.

PAGE 111

111 7.6.4 Tree Height Using Terrascan, points containe d in a cylindrical volume centered at the stem location coordinates were extracted from the “Forest Cube”; the maximum and minimum height points were obtained and the difference between them wa s computed yielding the tree height. Figure 712 shows a rendering of the points extracted for a si ngle tree with a height histogram used for the determination of the tree height. Figure 7-12. Single tree he ight determination.

PAGE 112

112 7.6.5 Stem Volume Tree stem volume may be a vague concept; it may be used to re fer to the total stem volume or to the “Merchantable Volume”, which is define d for a particular length of at trunk up to which a particular product may be obtained. There ar e many ways that foresters estimate the stem volume, and they vary greatly from place to place, based on the purpose and tree species. Usually foresters have tables that require tr ee species, diameter at breast height and the merchantable height (or total tree height) as inputs. The tables ar e based on allometric equations which are empirical regressions that relate stem volume and biomass of sp ecies to diameter at breast height and/or to tree height. For this particular dataset it is impossible to determine the stem merchantable volume because the trees on the plantation are relatively young and have not reached the diameter and height at which they are considerable exploitable. However the point cloud can be used to estimate the total stem volume (including bark) from the ground to the crown. TerraScan was employed to obtain the tree diameter at di fferent heights as shown in Figure 7-13. Figure 7-13. Diameters at different heights for volume computations

PAGE 113

113 Each section between diameter measurements can be approximated as a truncated cone from which the total stem volume was computed. Table 7-3 summarizes th e diameter and height measurements that were used for the stem volume computation. Table 7-3. Diameter and heights measurements for stem volume estimation Height above ground [m] Diameter [m] Area [m] Segment height Segment volume 0 0.2295 0.041367 1.4 0.2295 0.041367 1.4 0.057914 3.747 0.2275 0.040649 2.347 0.096245 5.933 0.1896 0.028234 2.186 0.074878 7.8004 0.1516 0.01805 1.8674 0.042862 9.2853 0.1453 0.016581 1.4849 0.025705 11.0672 0.1485 0.01732 1.7819 0.030202 12.7385 0.1295 0.013171 1.6713 0.025401 15.8157 0.079 0.004902 3.0772 0.02678 18.395 0 0 2.5793 0.004214 Total stem volume [m] 0.384201 Total stem volume modeled as cone regressed from data [m] 0.394114637 7.6.6 Tree Biomass Estimation There are multiple formulas that provide estimates of the total aboveground biomass for all hardwood and conifer species in the U.S. One fo rmula, presented by Jenkins et al. (2003) is: dbhe bmln1 0 where: bm = total aboveground biomass (kg) for trees 2.5-cm dbh and larger dbh. = diameter at breast height (cm) 0 and 1 are fit parameters for each species For the tree sample that has been worked thr oughout this chapter the result of the biomass estimate is: kg e e bmdbh99 16295 22 ln 4349 2 5356 2 ln1 0 7.7 Comparison with Traditional Methods Several of the individual tree metrics were obtai ned or derived from the traditional forestry techniques and methods as depicted in Figure 7-14.

PAGE 114

114 Figure 7-14. Individual tree metric measurement. A) Circumference at diameter height to determine DBH. B) Estimation of tree height. For the tree that was used to derive its metrics from the LIDAR point cloud, 6 measurements of its circumference were measured (62.5, 60.5, 61.2, 61.5, 61.0, 61.8 cm) which yielded and average DBH of 19.55 cm. Also three es timates of its height were obtained using the traditional Biltmore Stick (19.28, 17.36, 18.2 m) wh ich yields an average tree height of 18.28 m. From these measured metrics several others can be derived, a summary of the comparison of the directly measured and derived metrics obtained from the point cloud and traditional metrics are presented on Table 7-4. Table 7-4. Comparison of the metrics derived fr om traditional methods and from the TLS point cloud. Metric Traditional Method From M-TLS point cloud % Difference DBH 0.1955 m 0.2295 m 17.39 % Tree Height 18.28 m 18.395 m 0.63 % Stem Volume 0.3842 m Stem Volume approximated to Cone 0.1829 m 0.2536 m 7.07% Biomass 110.31 kg 162.99 kg 47.8 % 7.8 Conclusions It was possible to extract plot and individual tree metrics from the fused data set combining both airborne and terrestrial scanners. However th ere is more work to be done in the areas of merging and geo-referencing of the terrestrial po int clouds, such that a mo re coherent data set

PAGE 115

115 will be available for data extraction. All the work presented here was done in a manual fashion. However, all of the activities can be written into algorithms that wi ll automate the data extraction process. The synergy of airbor ne and terrestrial laser scanni ng technologies can prove highly valuable to Forest Science, providing vast amounts of info rmation that can enable the improvement of the allometric models.

PAGE 116

116 CHAPTER 8 SUMMARY 8.1 Conclusions Terrestrial Laser Scanners (TLS) are very ve rsatile measuring and ma pping tools, but as with any other tool, to become the “method of choi ce” they have to prove to be efficient in their application. This means that they have to be better, faster and cheap er than the alternate equipment and techniques. Better, faster and cheaper implies both qualitative and quantitative improvements; such as more measurements with gr eater accuracy at significant savings in time, labor and capital investments. In their short lif etime TLS units have prov en their advantage over traditional surveying equipment and methods in ma ny engineering applications such as As-Built documentation and transportation infrastructure ma nagement. With respect to the application of TLS to scientific mapping, measurement and documen tation; it is clear that they are faster and better than the traditional techniques. What is left is to prove is that they can also be more economic. For the tested applications in the fields of geology, paleontology, forestry and coastal morphology the ILRIS unit and/or the M-TLS system proved to collect data more accurately at higher resolutions and faster than the conventio nal techniques. The resu lting dataset that can easily exceed several hundreds of megabytes is both the TLS great est strength and weakness. On the positive side the dataset cont ains a wealth of spatial and spectral information from which an infinite amount of measurements can be derived, and the researcher can revisit the “virtual site” over and over again without leaving the office. On the negative side, more data is not necessarily better. In some cases more data can overwhel m the researcher. It can burry the phenomenon of interest under millions of unnecessary data points. In some extreme cases the analytical models used to describe the phenomenon simply can not handle the extreme resolu tion of a TLS dataset.

PAGE 117

117 Finally, many researchers are interested in obt aining their metrics and can be careless about performing all the preprocessing (p oint cloud cropping, merging and geo-referencing) required to have a usable dataset. However, these disadvantages or weakness can be capitalized into an opportunity to help develop str ong relationships between scien tists and geosensing engineers thru collaborative research projects. A result of all the tested a pplication was the investment of hundred of hours in data analysis which yielded a wealth of experience in the operation of LIDA R processing software. From that experience it can be concluded that th ere is no single software package that provides all possible tools required for th e application of TLS dataset for scientific purposes. For most of the projects operations were perf ormed in at least 3 software packages: Polyworks, Terrascan and QT Modeler; in some cases Surfer was employed to create grids and Matlab was used to code special customized codes for the extr action of the required final information. 8.2 Recommendations There is a need to develop methods to im prove the accuracy of the merging and georeferencing procedures. Extensive research shou ld be performed by future students on those fields as well as the establishment of a set of accuracy standards that will provide a benchmark for measuring the quality of these operations. It is also necessary to work in the field of sensor characteriza tion, which includes the development of test protocols to verify TLS specifications such as accuracy (under various ranges, target reflectance and scan angles), scan speed and measurement repeatability. This activity also includes the identification and charact erization of systematic errors for individual units. If a good test protocol is developed it will set the ground for an independent experiment to test and compare different makes and models of TLS units which can provide material for a great engineering paper.

PAGE 118

118 The development of the Mobile Terrestrial Laser Scanning sy stem must be accelerated, more students must be involved in the project and they can pa rticipate more actively in the integration of the different system subcomponents. One of the lessons learned from this work is that the multidisciplinary background of the students from both the GSE and AS PL groups is an invaluable academic and research asset. However more actions have to be taken toward s increasing the synergies between its members with respect to the application of MTLS to diverse scientific disciplines. Networking activities must be undertaken with in the UF academic community to identify potential partners that can use TLS systems in their research. Collaborative research projects should be started; the results can yield publishable material. Some consideration should be given to devel op a program such as NCALM that will make the M-TLS system and services available to scientific community. This program will not only increase system usage, but also increase the operator experience, the capability testing and provide more opportunities and material to publish. Efforts should be made to participate in TLS related conferences such as the International Society for Photogrammetry and Remote Se nsing Commission III on Photogrammetric Computer Vision and Image Anal ysis, where a great part of the published work refers to processing and interpretation of la ser range data. Also an annual meeting from an engineering and industry perspective is or ganized by Spar Point Research LLC; this meeting focuses on advanced dimensional control, work proce sses and 3D laser sca nning technologies. The participation in events of this sort will help UF researchers to keep on the leading edge of scientific and engineering applications of Terrestrial Laser Scanning systems.

PAGE 119

119 APPENDIX COMPARISON OF TERRESTRIAL LASER SCANNERS A.1 Optech ILRIS 3D This is a dual scanning mirror, pulsed infr ared laser system with the highest range available on the market: from 3 m to beyond 1 km. It has an integrated megapixel digital camera and large-format LCD viewfinder. 40 x 40 inst antaneous field of view. An optional Pan and Tilt base allows for a panoramic -20 to 90 x 360 scanning coverage. Web site Reference: http://www.optech.ca/i3dfeat-ilris.htm Figure A-1. Optech ILRIS TOF TLS A.2 Leica HDS3000 Leica High-Definition Surveying 3000, Nd:YAG frequency doubled pulsed laser, with a single mirror scanner, with a dua l-window and rotating base that allows for a 360 x 270 field of view. Has an integrated Bore-sighted Single 24 x 24, 1024 x 1024 digital camera for automatically calibrated photo ove rlays. Web site Reference: http://www.leicageosystems.com/hds/en/lgs_5574.htm Figure A-2. Leica HDS 3000 TOF TLS

PAGE 120

120 A.3 Leica HDS4500 25 & 53m The Leica HDS4500 is an ultra-high speed 690 nm laser, phase-based, short range scanner. Capable of scanning from 100,000 points/sec to 500,000 points/sec on a 360 x 310 field of view. Two different models are capable of 25 and 53 m ambiguous ranges. Web site Reference: http://www.leica-geosystems.com/hds/en/lgs_5572.htm Figure A.-3. Leica HDS 4500 PD TLS A.4 RIEGL LMS-Z420i Designed as a High-Accuracy & Long-Range pulsed laser, single vertical line rotating polygon scanner with a 80 range. The azimuth s can is accomplished by a rotating base full 360. An optional digital camera can be mounted for photorealistic textured point clouds. Range up to typ. 1000 m, precision up to 10 mm and scan rate up to 12 000 pts/s. Web site Reference: http://www.riegl.com/terrestrial _scanners/lms-z420i_/420i_all.htm Figure A-4. Riegl LMS-Z420i TOF TLS

PAGE 121

121 A.5 RIEGL LMS-Z390 Designed as a High-Accuracy & High Resolution pulsed laser, single vertical line rotating polygon scanner with a 80 range. The azimuth s can is accomplished by a rotating base full 360. An optional digital camera can be mounted for photorealistic textured point clouds. Range up to typ. 300 m, precision up to 2 mm and scan rate up to 12 000 pts/s. Web site Reference: http://www.riegl.com/terrestri al_scanners/lms-z390_/390_all.htm Figure A-5. Riegl LMS-Z390 TOF TLS A.6 RIEGL LMS-Z210ii Designed as a general purpose pulsed laser, single verti cal line rotating polygon scanner with a 80 range. The azimuth scan is accomp lished by a rotating base full 360. An optional digital camera can be mounted for photorealistic textured point clouds. Range up to typ. 650 m, precision up to 10 mm and scan rate up to 12 000 pts/s. Web site Reference: http://www.riegl.com/terrestrial_s canners/lms-z210ii_/210ii_all.htm Figure A-6. Riegl LMS-Z210ii TOF TLS

PAGE 122

122 A.7 Trimble GS101 A single scan line pulsed green laser system with an integrated digital camera, on a rotating base. It allows to scan in 60 vertical and 360 horizontal, up to a range of 200 -350 m, with a spot spacing of 32 rad and a range resolu tion of 1.5 mm @ 50m and a scan rate of up to 5,000 pts/s. Web site Reference: http://www.trimble.com/gs200.shtml Figure A-7. Trim ble GS101 TOF TLS A.8 Trimble GX 3D A single scan line pulsed green laser system with an integrated digital camera, on a rotating base. It allows to scan in 60 vertical and 360 horizontal, up to a range of 200 -350 m, with a spot spacing of 32 rad and a range resolu tion of 1.5 mm @ 50m and a scan rate of up to 5,000 pts/s. Web site Reference: http://www.trimble.com/trimblegx.shtml Figure A-8. Trimble GX 3D TOF TLS

PAGE 123

123 A.9 Minolta VIVID 910 The VI-910 uses LASER triangulation. The entire area is captured in 2.5 seconds, and the surface shape is converted to a lattice of ove r 300,000 vertices (connected points). A (24-bit) color image is captured at the same time by the same triangulation CCD. The range is limited to 2.5 meters and the range resolu tion is in the order or 8 micrometers. Web site Reference: http://www.minolta3d.com/products/main-en.asp Figure A-9. Minolta VIVID 910 OT TLS A.10 Zoller-Frohlish IMAGER 5006 Is an ultra-high speed visible laser, phase-ba sed, short range scanne r. Capable of scanning from up to 500,000 points/s on a 360 x 310 field of view with an Ambiguity range up to 79 meters. Web site Reference: http://www.zf-laser.com/e_imager5006.html Figure A-10. Zoller-Frohlish IMAGER 5006 PD TLS

PAGE 124

124 A.11 IQSun 880 Is an ultra-high speed visible 785 nm laser, ph ase-based, single line short range scanner. Capable of scanning from up to 240,000 points/ s on a 360 x 320 field of view with an Ambiguity range up to 76 meters. Figure A-11. IQSun 880 PD TLS A.12 Comparison of Terrestrial LASER Scanner Specifications The specifications of commercial TLS units was compiled and presented in Table A-1. Table A-1. Comparison of terrestrial laser scanner specifications Manufacturer Optech Leica Leica Leica Model ILRIS 3D HDS3000 HDS4500 25m HDS4500 53m Method of operation Time of Flight Time of Flight Phase Difference Phase Difference Range [m] 3-1500 @80% 3-800 @20% 3-350 @4% 300 @90% 134 @18% 0.1-25.2 0.1-53.5 Range resolution [mm] 4 4 3 5 Azimuth, elevation resolution [] 0.00115 0.00022918 20.0535228 20.0535228 IFOV [Ver X Hor] 40 x 40 Aux FOV [Ver X Hor] -20 to 90 x 360 270 x 360 310 x 360 310 x 360 Laser type/color Infrared Visible Green Visible Red Visible Red Laser wavelength [nm] 1500 532 690 690 Scan rate [points/s] 2,000 4,000 500,000 500,000 Beam divergence [] 0.00974 0.00687549 0.02864789 0.02864789 Texture Intensity & RGB Intensity & RGB Weight [kg] 12 17 Dimensions (LxWxH) [cm] 32 32 22 26.5 x 37 x 51 18 x 30 x 35 19 x 30 x 35 Power Supply 24 VDC 36 V, AC or DC 24V DC 90 260V AC 24V DC 90 260V AC Power consumption [W] 75 80 50-70 50-71

PAGE 125

125 Table A-1. Continued Manufacturer RIEGL RIEGL RIEGL Trimble Model LMS-Z420i LMS-Z390 LMS-Z210ii GS101 Method of operation Time of Flight Time of Flight Time of Flight Time of Flight Range [m] 2-1000 @80% 2-350 @10% 1-300 @80% 1-100 @10% 4-650 @80% 4-200 @10% 2-200 Range resolution [mm] 10? 6? 15? 1.4-6.5 Azimuth, elevation resolution [] 0.002, 0.0025 0.001 0.005 1.83346494 IFOV [Ver X Hor] 80 x 0.014 80 x 0.014 80 x 0.15 Aux FOV [Ver X Hor] 80 x 360 80 x 360 80 x 360 60 x 360 Laser type/color Near Infrared Near Infrar ed Near Infrared Green Laser wavelength [nm] Scan rate [points/s] 8,000-12,000 8,000-11,000 8,000-12,000 5,000 Beam divergence [] 0.014 0.014 0.15 0.00343775 Texture Intensity & RGB Intensity & RGB Intensity & RGB Intensity & RGB Weight [kg] 14.5 14.5 13 12.8 Dimensions (LxWxH) [cm] 21 x 21 x 46.3 21 x 21 x 46.3 21 x 21 x 43.7 34 x 27 x 42 Power supply 12-18 VDC 12-18 VDC 12-18 VDC 90-240 VAC 24 VDC Power consumption [W] 78-98 78-98 78-98 Table A-1. Continued Manufacturer Trimble Trimble Minolta Zoller-Frohlish Model GS200 GX 3D VIVID 910 IMAGER 5006 Method of operation Time of Flight Time of Flight Optical Triangulation Phase Difference Range [m] 2-350 350 @90% 0.6 -2.5 79 Range resolution [mm] 1.4-6.5 7 0.008 0.1 Azimuth, elevation resolution [] 0.00171887 0.00343775 0.0018 IFOV [Ver X Hor] Aux FOV [Ver X Hor] 60 x 360 60 x 360 310 x 360 Laser type/color Green Green Visible Red Visible Laser wavelength [nm] 532 690 Scan rate [points/s] 5,000 5,000 120,000-256,000 500,000 Beam divergence [] 0.00343775 0.00343775 0.01260507 Texture Intensity & RGB Intensity & RGB none Weight [kg] 12.8 13 11 14 Dimensions (LxWxH) [cm] 35 x 27 x 42 32.3 x 34.3 x 40.4 21.3 x 27.1 x 41.3 28.6 x 19 x 37.2 Power supply 90-240 VAC 24 VDC 90-240 VAC 24 VDC 100 to 240 VAC 24V DC Power consumption [W] 100 60 50

PAGE 126

126 LIST OF REFERENCES Baltsavias, E.(1999) Airborne la ser scanning: existing systems and firms and other resources. ISPRS Journal of Photogrammetry & Remote Sensing 54, 164–198. Baltsavias, E.P. (1999) Airborne laser scanning: basic relations and formulas. ISPRS Journal of Photogrammetry & Remote Sensing 54, 199–214. Bellian, J.A.; Jennette, D.C.; Kerans, C.; Gibeaut, J.; Andrews, J.; Yssldyk, B.; Larue, D. (2002) 3-Dimensional Digital Outcrop Data Collec tion and Analysis Using Eye-safe Laser (LIDAR) Technology. Published by AAPG/Datapa ges, Inc., 1444 South Boulder, Tulsa, OK, 74119, USA Retrieved March, 2007 from http://www.searchanddiscovery.com/documents/beg3d/index.htm Bender, Currie, Dickey, Eckhardt, Faller, Kaula, Mulholland, Plotkin, Poultney, Silverberg, Wilkinson, Williams, Alley. (1973). The Luna r Ranging Experiment. Science, New Series, 182, (4109), 229-238. Bennett, W.R., 2000. Background of an Inversio n: The First Gas Laser. IEEE Journal on Selected Topics In Quantum Electronics, 6 (6), 869-875. Beraldin, J-A; Blais, F; Rioux, M; Domey, J; Gonzo,L; De Nisi,F; Comper,F; Stoppa, D; Gottardi, M; Simoni,A.(2003) Optimized Position Sensors for Flying-Spot Active Triangulation Systems. Proceedings of the F ourth International Conference on 3-D Digital Imaging and Modeling (3DIM’ 03), October 06 10, 2003, Banff, Canada. 2003, 2836. Bureau of Beaches and Coastal Systems (BBCC ). Florida Department of Environmental Protection (2004) Monitoring Standards for Beac h Erosion Control Projects. Tallahassee, Florida. Retrieved March, 2007 from http://www.dep.state.fl.us/beaches/ Burkard, R.K. (1983). Geodesy for the Layman (Fourth Revision). Defense Mapping Agency Technical Report TR-80-003. St. Louis, Missouri. Callens, M., Verhoest, N.E.C. and Davidson, M.W. J.. (2006) Parameterization of TillageInduced Single-Scale Soil Roughness from 4-m Profiles. IEEE Transactions on Geoscience and Remote Sensing. 44 (4), 878 – 888. Carter, W. (1973). The Lunar Laser Ranging Pointing Problem. Unpublished doctoral dissertation, University of Arizona, Tucson. Carter, W. & Shrestha, R. (1997) Airborne laser swath mappi ng: Instant snapshots of our changing beaches. In Proceedings of the Four th International Conf erence Remote Sensing for Marine and Coastal Environments, Envir onmental Research Institute of Michigan (ERIM), Ann Arbor, Michigan. p. 298–307. Carter, W., Shrestha, R., Slatton, K.C., Cossio, T., Shrestha, K. (2005) Geodetic Laser Scanning: Refractive Optics Offer Wide Variety of S can Patterns. Poster at the American Geophysical Union, Fall Meeting 200 5, San Francisco, California.

PAGE 127

127 Cheves, M. (1999). Geodimeter—The First Name in EDM. Professional Surveyor, 19 (2). Cheves, M. (2002).In the Air, On the Ground: Laser-based Technology from Optech. Professional Surveyor, 22 (10). Curless, B.; Levoy, M. (1995). Better Optical Triangulation through Spacetime Analysis. Procedings of the Fifth Inte rnational Conference on Comput er Vision (ICCV 95), June 2023, 1995, Massachusetts Institute of Tec hnology, Cambridge, Massachusetts, USA., 987994. Davidson, M.W.J., Toan, T.L., Mattia, F., Satali no, G., Manninen, T., Borgeaud, M. (2000). On the Characterization of Agri cultural Soil Roughness for Radar Remote Sensing Studies. IEEE Transactions on Geoscience and Remote Sensing, 38 (2), 630-640. Defense Mapping Agency (DMA) (1991) Error Th eory as applied to Mapping Charting and Geodesy. DMA Technical Report 84 00.1, Fairax, Virginia, USA. Drake, D.R. (2002) Applications of Laser Scan ning and Imaging Systems. Unpublished Masters Thesis, University of Florida, Gainesville. Ewing, Clair & Mitchell, Mich ael (1970). Introduction to Ge odesy. Elsevier North Holland Publishing Company, Inc, New York, New York, USA. Flood, M. (2001) Laser Altimetry: From Scien ce to Commercial LIDAR Mapping. Journal of The American Society for Photogrammetr y and Remote Sensing, 67 (11), 1209-1211. Frhlich, C., Mettenleiter, M.(2004). Terrestri al Laser Scanning – New Perspectives In 3d Surveying. Proceedings of the ISPRS worki ng group 8/2 'Laser-Scanners for Forest and Landscape Assessment', Freiburg, Germ any, October 3-6, 2004, 7-13. Guenther, G., Goodman, L., Ena bnit, D., Swift, R., Thomas, R.. (1978). Laser Bathymetry for Near Shore Charting Applicati on (Preliminary Field Test Resu lts). Oceans, 10, 390396. Hickman, D., Hogg, J. (1969) Application Of An Airborne Pulsed Laser For Near Shore Bathymetric Measurements Remote Se nsing Of Environ, 1 (1), P 47-58. Hopkinson,C.; Chasmer, L.; Young-Pow, C.; Treitz P. (2004) Assessing fo rest metrics with a ground-based scanning LIDAR. Canadian J ournal of Forest Research. 34, 573–583. Iavarone, A (2002) Laser Scanner Fundamentals. Professional Surveyor Magazine, 24 (5). Institute of Electrical and Elect ronics Engineers (IEEE). (N.D.). Theodore H. Maiman. Published by the IEEE Virtual Museum, New Brunswick, New Jersey, USA. Retrieved March, 2007, from http://www.ieee-virtual-museum.org/c ollection/people.php?id=1234591&lid=1 Institute of Electrical and Elect ronics Engineers (IEEE). (N.D.). Charles H. Townes. Published by the IEEE Virtual Museum, New Brunswick New Jersey, USA.Retrieved March, 2007, from http://www.ieee-virtual-museum.org/c ollection/people.php?id=1234588&lid=1

PAGE 128

128 Institute of Electrical and Elect ronics Engineers (IEEE). (N.D.). Arthur L. Schawlow. Published by the IEEE Virtual Museum, New Brunswick New Jersey, USA.Retrieved March, 2007, from http://www.ieee-virtual-museum.org/c ollection/people.php?id=1234585&lid=1 Javan,A., Bennett, W.R., Herrott, D.R. (1961) Po pulation inversion and continuous optical maser oscillation in a gas discharge containing a He–Ne mixture. Phys. Rev. Lett., 6, 106–110. Jenkins, J.C.; Chojnacky, D.C.; Heath, L.S.; Birdsey, R.A. 2003. National-scale biomass estimation for United States tree species. Forest Science. 49(1), 12-35. Kern, J. (2001) Mapping Groud Zero. Profe ssional Surveyor, November 2001, 21 (10) Kim, H; Cervenka, P; Lankford, C. (1975) De velopment of an Airborne Laser Bathymeter, NASA Techincal Note D-8079, NAS A Wallops Flight Center, Wallops Island, Virginia, USA. Kraus, N.C. & Galgano, F.A. (2001) Beach Eros ional Hot Spots: Types, Causes, and Solutions. US Army Engineer Research and Developmen t Center/Coastal Hydraulics Laboratory. Coastal and Hydraulics Engineering T echnical Note ERDC/CHL CHETN-II-44. Vicksburg, Mississippi, USA. Kaufman, C.C., Batts, E.L., Levine, N.S., D oyle, B.C. (2005) Beach Erosion Mapping Using LIDAR and a 3D Laser Scanner: Folly Beach, South Carolina. Poster presentation at the Geological Society of Amer ica (GSA) Annual Meeting 200 5, Salt Lake City, Utah, October 16–19, 2005, paper 59-14 Lefsky,M.; Cohen, W.; Parker, W.; & Hardi ng, D. (2002) LIDAR Remote Sensing for Ecosystem Studies. BioScience, 52 (1), 19-30. Lefsky, M.A.; Cohen, W.B.; Acker, S. A.; Spies, T.A.; Parke3, G.G.; Harding, D. (1998), LIDAR Remote Sensing of Forest Canopy Stru cture and Related Biophysical Parameters at the H.J. Andrews Experimental Forest, Oregon, USA. Geoscien ce and Remote Sensing Symposium Proceedings, 1998. IGARSS '98, Seattle, WA, USA. 6-10 Jul 1998, 3, 12521254. Lichti, D.D.; Gordon, S.J. & Stewart, M.P .. (2002). Ground-based lase r scanners: operation, systems and applications. Geomatica, 56, 22–33. Maling, D. H. (1992). Coordina te Systems and Map Projectio ns (2nd Edition). Pergamon, Oxford, United Kindom. Marino, R.; Davis, W.R. (2005) Jigsaw: A Folia ge-Penetrating 3D Imaging Laser Radar System. Lincoln Laboratory Journal, 15 (1), 23-36. Martel, S.J. (2006) Effect of topographic curvature on near-surf ace stresses and application to sheeting joints: Geophysical Resear ch Letters, 33 (1) pg:L01308.

PAGE 129

129 Mattia, F.; Davidson, Malcolm W.J.; Toan, Thuy Le ; D’Haese, Christophe M. F.; Verhoest, Niko E.C.; Gatti, Anna Maria; Borgeaud, Ma urice. (2003) A Comparison Between Soil Roughness Statistics Used in Surface Scattering Models De rived From Mechanical and Laser Profilers. IEEE Transactions on Geosci ence and Remote Sensing, 41 (7), 1659-1671. McClung, F.J. and Hellwarth, R.W. (1962): "Giant optical pulsations from ruby". Journal of Applied Physics, 33 (3), 828-829. McGarry, Jan; Zagwodzki, Tom. (2005) A Brie f History of Satellit e Laser Ranging: 1964 – present*. Published by the Crustal Dynamics Data Information System (CDDIS), NASA Goddard Space Flight Center, Greenbelt, Maryland. Retrieved October, 2005, from http://cddis.nasa.gov/slr2000/docs/gsfcslr_jm0504.pdf Mitchell, R (N.D.) AOL History WFF AOL. Published by the Hydros pheric and Biospheric Sciences Laboratory, NASA GSFC/Wallops Flight Facility, Wallops Island, Virginia, USA Retrieved March, 2007, from http://aol.wff.nasa.gov/index.php?modul e=pagemaster&PAGE_user_op=view_page&PAG E_id=3&MMN_position=3:3 Nsset, E & Gobakken, T. (2005) Estimating fore st growth using canopy metrics derived from airborne laser scanner data. Remote Sensing of Environment 96, 453 – 465. National Aeronautics and Space Administration (NASA). (N.D.) National Space Science Data Center Apollo Laser Altimeter Experiment Published by the NASA Goddard Space Flight Center, Greenbelt, Maryland. Retrieved March, 2007, from http://nssdc.gsfc.nasa.gov/databa se/MasterCatalog?sc=1971-063A&ex=5 National Oceanic and Atmospheric Administration (NOAA), (N.D.). Distance Measurement Tools, AGA Geodimeter NASM-2A. Published by the NOAA the Office of Public Affairs, Washington, D.C., USA. Re trieved March, 2007, from http://celebrating200years.no aa.gov/distance_tools/0411.html National Oceanic and Atmospheric Administration (NOAA), (N.D.). Distance Measurement Tools, Tellurometer. Published by the NOAA the Office of Public Affairs, Washington, D.C., USA.Retrieved March, 2007, from http://celebrating200years.noaa.gov/dist ance_tools/tellurometermra1.html National Oceanic and Atmospheric Administration (NOAA), (N.D.). Distance Measurement Tools, Ranger III and Rangemaster III. P ublished by the NOAA the Office of Public Affairs, Washington, D.C., US A.Retrieved March, 2007, from http://celebrating200years.noaa. gov/distance_tools/ranger.html Oelze, Michael L.; Sabatier, James M.; Rasp et, Richard (2003). Roughness Measurements of Soil Surfaces by Acoustic Backscatter. Soil Scie nce Society of America Journal, 67, 241250.

PAGE 130

130 Parrish, C.E.; Tuell, G.H.; Carter,W.E. and Shrest ha, R.L. (2005) Configuring an Airborne Laser Scanner for Detecting Airport Obstructions Photogrammetric Engineering & Remote Sensing, 71 (1), 37–46. Perron, Jay Taylor (2006). Formation of Evenly Spaced Ridges and Valleys Unpublished doctoral dissertation, Universi ty of California, Berkeley. Schawlow, A. L. & Townes, C. H. (1958) Infrared and Optical Masers. Physical Review, 112, 1940-1949. Shrestha, Ramesh (1983). Local Geodetical Networks Adjustments in Three Dimensions Unpublished doctoral dissertation, Un iversity of Wisconsin-Madison. Shrestha,R.L., Carter,W.E., Sartori,M., Luzum, B.J., Slatton, K.C.(2005) Airborne Laser Swath Mapping: Quantifying changes in sandy beaches ove r time scales of weeks to years. ISPRS Journal of Photogrammetry & Remote Sensing, 59, 222–232. Smith, James R. (1996). Introduction to Geodesy John Wiley & Sons, Inc. New York, New York, USA. Smithsonian National Museum of American Histor y. Virtual Surveying Instrument Collection. (ND). EDM (Geodimeter Model 2A). Publishe d by the Smithsonian National Museum of American History. Washington. D.C ., USA. Retrieved March, 2007, from http://americanhistory.si.edu/collections /surveying/object.cfm?recordnumber=748815 Smithsonian National Museum of American Histor y. Virtual Surveying Instrument Collection. (ND). EDM (Tellurometer M/RA 1). Published by the Smithsonian National Museum of American History. Washington. D.C ., USA. Retrieved March, 2007, from http://americanhistory.si.edu/collections /surveying/object.cfm?recordnumber=759173 Smithsonian National Museum of American Histor y. Virtual Surveying Instrument Collection. (ND). EDM (Tellurometer M/RA 1). Published by the Smithsonian National Museum of American History. Washington. D.C ., USA. Retrieved March, 2007, from http://americanhistory.si.edu/collecti ons/surveying/maker.cfm?makerid=45 Smithsonian National Museum of American Histor y. Virtual Surveying Instrument Collection. (ND). EDM (Geodolite). Published by the Sm ithsonian National Museum of American History. Washington. D.C., USA. Retrieved March, 2007, from http://americanhistory.si.edu/collections /surveying/object.cfm?recordnumber=748473 Taylor, Nick (2000). LASER: The inventor, the Nobel laureate, a nd the thirty-year patent war. Simon & Schuster. New York, New York, USA. Thoma,D. P., Moran,M.S., Bryant,R., Rahman,M., Holifield-Collins,C.D., Sano,E.E., Slocum, K., Skirvin,S..(2006) Comparison of four mode ls to determine surface soil moisture from C-band radar imagery in a sparsely vegeta ted semiarid landscape. Water Resources Research, Vol. 42, W01418

PAGE 131

131 Watt, P. J. & Donoghue, D. N. M. (2005) Measur ing forest structure w ith terrestrial laser scanning. International Journal of Remote Sensing, 26 (7), 1437–1446. Wehr, A; Lohr U. (1999) Airborne Laser S canning—An Introduction and Overview. ISPRS Journal of Photogrammetry & Remote Sensing 54, 68–82. Zribi,M., Ciarletti, V., Tac onet, O., Boissard, P., Chapr on, M. and Rabin, B. (2000) Backscattering on soil structure described by plane facets. International Journal on Remote Sensing, 21 (1), 137-153.

PAGE 132

132 BIOGRAPHICAL SKETCH Juan Carlos Fernandez Diaz was born Augus t 18, 1976, in Tegucigalpa, Honduras, to Venancio Fernandez and Ana Maria Diaz. His fam ily is also composed by one sister: Maria Esther and two brothers David and Jose Venancio. Since a very young age he developed strong interest towards science and technology; espe cially to earth and sp ace science, aviation, telecommunications and electroni cs. He was fortunate to atte nd an American school (Elvel School) from kindergarten to 11th grade, where the professors motivated his scientific curiosity. He graduated from High School in 1993 from a program that fulfills the requirements from both the American and Honduran academic curriculum. Th at same year he enrolled on the Electrical Engineering program of the Universidad N acional Autonoma de Hondur as (UNAH). When he did not find college challenging e nough he decided to work full time while pursuing the B.S. degree. His first position was as an instructor of the university’ s astronomical observatory where he acquired knowledge and expertis e related to the desi gn, use and maintenance of astronomical instrumentation as well as astronomical data proc essing and analysis. During this period he also participated in a train ship at the European Space Agency Satellite Tracking Station in Villafranca del Castillo, Spain and received a Summer Undergraduate Research Fellowship (SURF) from the California Inst itute of Technology (CALTECH) to perform scientific research at the Jet Propulsion Laboratory. Soon after this experience he accepted a new position with the Honduras National Telecommunica tion Commission as a Spectru m Planning and Engineering technician. He obtained the BS degree in electrical engineering in June 2001 having not only the formal academic knowledge but also a great de al of experience in telecommunications, space science and technology. From 2002 to 2005 he continued his career in telecommunications holding positions in a Wireless service provider were he performed func tions as network planni ng engineer and quality

PAGE 133

133 assurance chief. He also obtai ned a Master of business admini stration degree with a summa cum laude distinction from the Universidad Cato lica de Honduras in 2005. During 2004 he applied for a Fulbright Scholarship to participate in a masters program in the fields of Satellite Applications (Navigation, Comm unications and Remote Sensing). He was fortunate to receive the scholarship and to be accepted to the Univers ity of Florida, Geosensing Systems Engineering graduate program. Juan hopes to continue with his multidiscip linary education, continue to explore his interest in space science and t echnology and to contribute in some way to mankind progress, especially to the benef it of his fellow Hondureos.