Publicação

Digital Multispectral Map Reconstruction Using Aerial Imagery

Ver documento

Detalhes bibliográficos
Resumo:Advances made in the computer vision field allowed for the establishment of faster and more accurate photogrammetry techniques. Structure from Motion(SfM) is a photogrammetric technique focused on the digital spatial reconstruction of objects based on a sequence of images. The benefit of Unmanned Aerial Vehicle (UAV) platforms allowed the ability to acquire high fidelity imagery intended for environmental mapping. This way, UAV platforms became a heavily adopted method of survey. The combination of SfM and the recent improvements of Unmanned Aerial Vehicle (UAV) platforms granted greater flexibility and applicability, opening a new path for a new remote sensing technique aimed to replace more traditional and laborious approaches often associated with high monetary costs. The continued development of digital reconstruction software and advances in the field of computer processing allowed for a more affordable and higher resolution solution when compared to the traditional methods. The present work proposed a digital reconstruction algorithm based on images taken by a UAV platform inspired by the work made available by the open-source project OpenDroneMap. The aerial images are inserted in the computer vision program and several operations are applied to them, including detection and matching of features, point cloud reconstruction, meshing, and texturing, which results in a final product that represents the surveyed site. Additionally, from the study, it was concluded that an implementation which addresses the processing of thermal images was not integrated in the works of OpenDroneMap. By this point, their work was altered to allow for the reconstruction of thermal maps without sacrificing the resolution of the final model. Standard methods to process thermal images required a larger image footprint (or area of ground capture in a frame), the reason for this is that these types of images lack the presence of invariable features and by increasing the image’s footprint, the number of features present in each frame also rises. However, this method of image capture results in a lower resolution of the final product. The algorithm was developed using open-source libraries. In order to validate the obtained results, this model was compared to data obtained from commercial products, like Pix4D. Furthermore, due to circumstances brought about by the current pandemic, it was not possible to conduct a field study for the comparison and assessment of our results, as such the validation of the models was performed by verifying if the geographic location of the model was performed correctly and by visually assessing the generated maps.
Autores principais:Vong, André Agostinho Cheang do Rosário
Assunto:Remote Sensing Photogrammetry Computer Vision UAV Structure from Motion Digital Reconstruction
Ano:2021
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso aberto
Instituição associada:Universidade Nova de Lisboa
Idioma:inglês
Origem:Repositório Institucional da UNL
_version_ 1868982991313698816
author Vong, André Agostinho Cheang do Rosário
author_facet Vong, André Agostinho Cheang do Rosário
author_role author
contributor_name_str_mv Mora, André
Carvalho, João
RUN
country_str PT
creators_json_txt [{\"Person.name\":\"Vong, André Agostinho Cheang do Rosário\"}]
datacite.contributors.contributor.contributorName.fl_str_mv Mora, André
Carvalho, João
RUN
datacite.creators.creator.creatorName.fl_str_mv Vong, André Agostinho Cheang do Rosário
datacite.date.Accepted.fl_str_mv 2021-12-01T00:00:00Z
datacite.date.available.fl_str_mv 2022-02-21T14:53:03Z
datacite.date.embargoed.fl_str_mv 2022-02-21T14:53:03Z
datacite.rights.fl_str_mv http://purl.org/coar/access_right/c_abf2
datacite.subjects.subject.fl_str_mv Remote Sensing
Photogrammetry
Computer Vision
UAV
Structure from Motion
Digital Reconstruction
datacite.titles.title.fl_str_mv Digital Multispectral Map Reconstruction Using Aerial Imagery
dc.contributor.none.fl_str_mv Mora, André
Carvalho, João
RUN
dc.creator.none.fl_str_mv Vong, André Agostinho Cheang do Rosário
dc.date.Accepted.fl_str_mv 2021-12-01T00:00:00Z
dc.date.available.fl_str_mv 2022-02-21T14:53:03Z
dc.date.embargoed.fl_str_mv 2022-02-21T14:53:03Z
dc.format.none.fl_str_mv application/pdf
dc.identifier.none.fl_str_mv http://hdl.handle.net/10362/133295
dc.language.none.fl_str_mv eng
dc.rights.none.fl_str_mv http://purl.org/coar/access_right/c_abf2
dc.subject.none.fl_str_mv Remote Sensing
Photogrammetry
Computer Vision
UAV
Structure from Motion
Digital Reconstruction
dc.title.fl_str_mv Digital Multispectral Map Reconstruction Using Aerial Imagery
dc.type.none.fl_str_mv http://purl.org/coar/resource_type/c_bdcc
description Advances made in the computer vision field allowed for the establishment of faster and more accurate photogrammetry techniques. Structure from Motion(SfM) is a photogrammetric technique focused on the digital spatial reconstruction of objects based on a sequence of images. The benefit of Unmanned Aerial Vehicle (UAV) platforms allowed the ability to acquire high fidelity imagery intended for environmental mapping. This way, UAV platforms became a heavily adopted method of survey. The combination of SfM and the recent improvements of Unmanned Aerial Vehicle (UAV) platforms granted greater flexibility and applicability, opening a new path for a new remote sensing technique aimed to replace more traditional and laborious approaches often associated with high monetary costs. The continued development of digital reconstruction software and advances in the field of computer processing allowed for a more affordable and higher resolution solution when compared to the traditional methods. The present work proposed a digital reconstruction algorithm based on images taken by a UAV platform inspired by the work made available by the open-source project OpenDroneMap. The aerial images are inserted in the computer vision program and several operations are applied to them, including detection and matching of features, point cloud reconstruction, meshing, and texturing, which results in a final product that represents the surveyed site. Additionally, from the study, it was concluded that an implementation which addresses the processing of thermal images was not integrated in the works of OpenDroneMap. By this point, their work was altered to allow for the reconstruction of thermal maps without sacrificing the resolution of the final model. Standard methods to process thermal images required a larger image footprint (or area of ground capture in a frame), the reason for this is that these types of images lack the presence of invariable features and by increasing the image’s footprint, the number of features present in each frame also rises. However, this method of image capture results in a lower resolution of the final product. The algorithm was developed using open-source libraries. In order to validate the obtained results, this model was compared to data obtained from commercial products, like Pix4D. Furthermore, due to circumstances brought about by the current pandemic, it was not possible to conduct a field study for the comparison and assessment of our results, as such the validation of the models was performed by verifying if the geographic location of the model was performed correctly and by visually assessing the generated maps.
dirty 0
eu_rights_str_mv openAccess
format masterThesis
fulltext.url.fl_str_mv https://run.unl.pt/bitstreams/790e1172-2d62-47ad-aeb5-c0e501644f10/download
funder_facet_str_mv FCT{{{_:::_}}}Fundação para a Ciência e a Tecnologia
FCT{{{_:::_}}}Fundação para a Ciência e a Tecnologia
funding.funder.alternateName_str_mv FCT
FCT
funding.funder.identifier_str_mv http://doi.org/10.13039/501100001871
http://doi.org/10.13039/501100001871
funding.funder.name_str_mv Fundação para a Ciência e a Tecnologia
Fundação para a Ciência e a Tecnologia
funding.name_str_mv 6817 - DCRRNI ID
6817 - DCRRNI ID
id run_4f9c0ddb2bba8cfcaa2d64d44830fdbd
identifier.url.fl_str_mv http://hdl.handle.net/10362/133295
inst_facet_str urn:organizationAcronym:unl{{{_:::_}}}Universidade Nova de Lisboa
instacron_str unl
institution Universidade Nova de Lisboa
instname_str Universidade Nova de Lisboa
language eng
network_acronym_str run
network_name_str Repositório Institucional da UNL
oai_identifier_str oai:run.unl.pt:10362/133295
organization_str_mv urn:organizationAcronym:unl
person_str_mv Vong, André Agostinho Cheang do Rosário
publishDate 2021
repo_facet_str urn:repositoryAcronym:run{{{_:::_}}}Repositório Institucional da UNL
reponame_str Repositório Institucional da UNL
repository_id_str urn:repositoryAcronym:run
service_str_mv urn:repositoryAcronym:run
spelling engpt_PTAdvances made in the computer vision field allowed for the establishment of faster and more accurate photogrammetry techniques. Structure from Motion(SfM) is a photogrammetric technique focused on the digital spatial reconstruction of objects based on a sequence of images. The benefit of Unmanned Aerial Vehicle (UAV) platforms allowed the ability to acquire high fidelity imagery intended for environmental mapping. This way, UAV platforms became a heavily adopted method of survey. The combination of SfM and the recent improvements of Unmanned Aerial Vehicle (UAV) platforms granted greater flexibility and applicability, opening a new path for a new remote sensing technique aimed to replace more traditional and laborious approaches often associated with high monetary costs. The continued development of digital reconstruction software and advances in the field of computer processing allowed for a more affordable and higher resolution solution when compared to the traditional methods. The present work proposed a digital reconstruction algorithm based on images taken by a UAV platform inspired by the work made available by the open-source project OpenDroneMap. The aerial images are inserted in the computer vision program and several operations are applied to them, including detection and matching of features, point cloud reconstruction, meshing, and texturing, which results in a final product that represents the surveyed site. Additionally, from the study, it was concluded that an implementation which addresses the processing of thermal images was not integrated in the works of OpenDroneMap. By this point, their work was altered to allow for the reconstruction of thermal maps without sacrificing the resolution of the final model. Standard methods to process thermal images required a larger image footprint (or area of ground capture in a frame), the reason for this is that these types of images lack the presence of invariable features and by increasing the image’s footprint, the number of features present in each frame also rises. However, this method of image capture results in a lower resolution of the final product. The algorithm was developed using open-source libraries. In order to validate the obtained results, this model was compared to data obtained from commercial products, like Pix4D. Furthermore, due to circumstances brought about by the current pandemic, it was not possible to conduct a field study for the comparison and assessment of our results, as such the validation of the models was performed by verifying if the geographic location of the model was performed correctly and by visually assessing the generated maps.application/pdfpt_PTDigital Multispectral Map Reconstruction Using Aerial ImageryVong, André Agostinho Cheang do RosárioMora, AndréCarvalho, JoãoHostingInstitutionOrganizationalRUNe-mailmailto:run@unl.ptrun@unl.pt2022-02-21T14:53:03Z2021-122021-12-01T00:00:00ZHandlehttp://hdl.handle.net/10362/133295http://purl.org/coar/access_right/c_abf2open accessRemote SensingPhotogrammetryComputer VisionUAVStructure from MotionDigital Reconstruction13958813 bytesFundação para a Ciência e a TecnologiaCOPELABS - Cognitive and People-centric Computing R&D Unit6817 - DCRRNI IDCrossref Funder IDhttp://doi.org/10.13039/501100001871Fundação para a Ciência e a TecnologiaCentre of Technology and Systems6817 - DCRRNI IDCrossref Funder IDhttp://doi.org/10.13039/501100001871literaturehttp://purl.org/coar/resource_type/c_bdccmaster thesishttp://purl.org/coar/access_right/c_abf2application/pdffulltexthttps://run.unl.pt/bitstreams/790e1172-2d62-47ad-aeb5-c0e501644f10/download
spellingShingle Digital Multispectral Map Reconstruction Using Aerial Imagery
Vong, André Agostinho Cheang do Rosário
Remote Sensing
Photogrammetry
Computer Vision
UAV
Structure from Motion
Digital Reconstruction
status SINGLETON
subject.fl_str_mv Remote Sensing
Photogrammetry
Computer Vision
UAV
Structure from Motion
Digital Reconstruction
title Digital Multispectral Map Reconstruction Using Aerial Imagery
title_full Digital Multispectral Map Reconstruction Using Aerial Imagery
title_fullStr Digital Multispectral Map Reconstruction Using Aerial Imagery
title_full_unstemmed Digital Multispectral Map Reconstruction Using Aerial Imagery
title_short Digital Multispectral Map Reconstruction Using Aerial Imagery
title_sort Digital Multispectral Map Reconstruction Using Aerial Imagery
topic Remote Sensing
Photogrammetry
Computer Vision
UAV
Structure from Motion
Digital Reconstruction
topic_facet Remote Sensing
Photogrammetry
Computer Vision
UAV
Structure from Motion
Digital Reconstruction
url http://hdl.handle.net/10362/133295
visible 1