000 05022nas a2200481 a 4500
005 20260817170900.0
008 190614s2011 th ad ||| 000 0 eng d
035 _a.b12133991
099 9 _aAIT Diss. no.RS-11-05
100 1 _aBandara, K. R. M. U.
245 1 0 _aAutomate generation of DEM from DSM in Forest Area using artificial neural networks
260 _aPathum Thani :
_bAsian Institute of Technology,
_c2011
300 _a1 online resource (51 leaves) :
_bill., charts
490 1 _aDissertation ;
_vno. RS-11-05
500 _aA dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Remote Sensing and Geographic Information Systems
502 _aThesis (Ph.D.) - Asian Institute of Technology, 2011
520 _aThisresearch was carried out touse the Digital Surface Model (DSM) to obtain a Digital Terrain Model (DEM) in a forest area by using Artificial Neural Networks (ANN) instead of removing DSM point clouds and then interpolation the formed hollow area using surrounding DEMpoints. The study was carried out for one of the dense forest areas ofMeegahakivula in Badulla District, Sri Lanka wherethe elevation is varying from 230 m to 403 m with approximately 16 m mean heights of trees. Initially, onlythe perimeter DEMdata was used to train the ANN and secondly, the training was done using only nine well distributed DEMpoints inside the forest area. An application oriented ANN software module was designed with the facilities to use all relevant parameters and it can be used to trainand use for any coordinate system transformations. The developed application oriented ANN was trained by using sample data. It was trained to 1.5 m accuracy for 172 points of the perimeter and 3.9 m for the nine points. By using trained ANN, the whole DSMdata of the area and the extracted DSM with the range of [min, max] and [mean {u2013}standard deviation, mean + standard deviation] of the used sample data sets were projected to DEM. The above projected DEMwith the two ranges were interpolated using severalinterpolation methods. All ANN projected DEMs were compared with the reference DEM to check the fitness of the ANN projected DEMsand obtained the RMSE as the deviation of projected DEMs with reference DEM. The highest accuracy as the lowest overall deviation (lowest RMSE)was obtained by the [min, max] range DSM projection with the nine points trained ANN and by Topo to Raster interpolation method and it was 1.240 m. Other DEMaccuracies were also closer to this and it was proven that the ability to project DSM to DEMwith ANN.DEM projected by ANN can be used to slope determination, landslide monitoring, soil erosion monitoring, orthophoto generation, eliminate of foreshortening and layover of SAR images, etc. according to the accuracy obtained presently, but has to be validated with different forest areas as well as all other areas separately. Further, the developed program has to be improved for mini-batch processing as well as to use for other applications.
650 0 _aRemote-sensing images
650 0 _aForests and forestry
_xRemote sensing
700 0 _aLal Lamarakoon,
_eChairperson
700 1 _akamiya, Yoshikazu,
_eExamination Committee
700 1 _aShrestha, Rajendra Prasad,
_eExamination Committee
710 2 _aGeoinformatics Center, Asian Institute of Technology,
_eScholarship Donor
710 2 _aNorth {u2013}South Center, Zurich, Switzerland,
_eScholarship Donor
710 2 _aInstitute of Photogrammetry & Remote Sensing, ETH, Hoenggerburg, Zurich, Switzerland,
_eScholarship Donor
810 2 _aAsian Institute of Technology.
_tDissertation ;
_vno. RS-11-05
856 _3Full-Text
_uhttp://203.159.5.9/ait-thesis/detail.php?q=B00757
907 _a.b12133991
_bmnait
_cj
902 _a240404
998 _b0
_c140619
_dm
_ea
_fj
_g0
962 _a000:001:PDF:b1213399:001549:0:0:0:0:0:0
_tAbstract--AIT Diss. no.RS-11-05
_vn
945 _lmnait
945 _lmnait
945 _lmnarc
945 _lmnarc
942 _c20
942 _c40
909 _aBarcode : 30050120555858
_bCREATED : 2014-06-19
_cRECORD # : i12813333
_dLPATRON : 0
_eLCHKIN : -
_f# RENEWALS : 0
_g# OVERDUE : 0
_hIUSE3 : 0
_iTOT CHKOUT : 0
_jTOT RENEW : 0
909 _aBarcode : 30050120555825
_bCREATED : 2014-06-19
_cRECORD # : i12813345
_dLPATRON : 1029963
_eLCHKIN : 2020-01-31
_f# RENEWALS : 0
_g# OVERDUE : 0
_hIUSE3 : 1
_iTOT CHKOUT : 1
_jTOT RENEW : 1
909 _aBarcode : 30050160000963
_bCREATED : 2016-03-23
_cRECORD # : i12902809
_dLPATRON : 0
_eLCHKIN : -
_f# RENEWALS : 0
_g# OVERDUE : 0
_hIUSE3 : 0
_iTOT CHKOUT : 0
_jTOT RENEW : 0
909 _aBarcode : 30050120993208
_bCREATED : 2020-12-02
_cRECORD # : i13259489
_dLPATRON : 0
_eLCHKIN : -
_f# RENEWALS : 0
_g# OVERDUE : 0
_hIUSE3 : 0
_iTOT CHKOUT : 0
_jTOT RENEW : 0
999 _c10111
_d10111