Multi-class strategies for joint building footprint and road detection in remote sensing

Building footprints and road networks are important inputs for a great deal of services. For instance, building maps are useful for urban planning, whereas road maps are essential for disaster response services. Traditionally, building and road maps are manually generated by remote sensing experts o...

ver descrição completa

Detalhes bibliográficos
Autores: Ayala Lauroba, Christian, Aranda, Carlos, Galar Idoate, Mikel
Formato: artículo
Estado:Versión publicada
Fecha de publicación:2021
País:España
Recursos:Universidad Pública de Navarra
Repositorio:Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
OAI Identifier:oai:academica-e.unavarra.es:2454/41735
Acesso em linha:https://hdl.handle.net/2454/41735
Access Level:acceso abierto
Palavra-chave:Sentinel-1
Sentinel-2
Remote sensing
Building detection
Road detection
Deep learning
Convolutional neural networks
Multi-class semantic segmentation
Binary semantic segmentation
Multi-task semantic segmentation
id ES_b68ddcee21303dbfc351e98bcf86943f
oai_identifier_str oai:academica-e.unavarra.es:2454/41735
network_acronym_str ES
network_name_str España
repository_id_str
dc.title.none.fl_str_mv Multi-class strategies for joint building footprint and road detection in remote sensing
title Multi-class strategies for joint building footprint and road detection in remote sensing
spellingShingle Multi-class strategies for joint building footprint and road detection in remote sensing
Ayala Lauroba, Christian
Sentinel-1
Sentinel-2
Remote sensing
Building detection
Road detection
Deep learning
Convolutional neural networks
Multi-class semantic segmentation
Binary semantic segmentation
Multi-task semantic segmentation
title_short Multi-class strategies for joint building footprint and road detection in remote sensing
title_full Multi-class strategies for joint building footprint and road detection in remote sensing
title_fullStr Multi-class strategies for joint building footprint and road detection in remote sensing
title_full_unstemmed Multi-class strategies for joint building footprint and road detection in remote sensing
title_sort Multi-class strategies for joint building footprint and road detection in remote sensing
dc.creator.none.fl_str_mv Ayala Lauroba, Christian
Aranda, Carlos
Galar Idoate, Mikel
author Ayala Lauroba, Christian
author_facet Ayala Lauroba, Christian
Aranda, Carlos
Galar Idoate, Mikel
author_role author
author2 Aranda, Carlos
Galar Idoate, Mikel
author2_role author
author
dc.contributor.none.fl_str_mv Institute of Smart Cities - ISC
Gobierno de Navarra / Nafarroako Gobernua, 0011-1408-2020-000008
dc.subject.none.fl_str_mv Sentinel-1
Sentinel-2
Remote sensing
Building detection
Road detection
Deep learning
Convolutional neural networks
Multi-class semantic segmentation
Binary semantic segmentation
Multi-task semantic segmentation
topic Sentinel-1
Sentinel-2
Remote sensing
Building detection
Road detection
Deep learning
Convolutional neural networks
Multi-class semantic segmentation
Binary semantic segmentation
Multi-task semantic segmentation
description Building footprints and road networks are important inputs for a great deal of services. For instance, building maps are useful for urban planning, whereas road maps are essential for disaster response services. Traditionally, building and road maps are manually generated by remote sensing experts or land surveying, occasionally assisted by semi-automatic tools. In the last decade, deep learning-based approaches have demonstrated their capabilities to extract these elements automatically and accurately from remote sensing imagery. The building footprint and road network detection problem can be considered a multi-class semantic segmentation task, that is, a single model performs a pixel-wise classification on multiple classes, optimizing the overall performance. However, depending on the spatial resolution of the imagery used, both classes may coexist within the same pixel, drastically reducing their separability. In this regard, binary decomposition techniques, which have been widely studied in the machine learning literature, are proved useful for addressing multiclass problems. Accordingly, the multi-class problem can be split into multiple binary semantic segmentation sub-problems, specializing different models for each class. Nevertheless, in these cases, an aggregation step is required to obtain the final output labels. Additionally, other novel approaches, such as multi-task learning, may come in handy to further increase the performance of the binary semantic segmentation models. Since there is no certainty as to which strategy should be carried out to accurately tackle a multi-class remote sensing semantic segmentation problem, this paper performs an in-depth study to shed light on the issue. For this purpose, open-access Sentinel-1 and Sentinel-2 imagery (at 10 m) are considered for extracting buildings and roads, making use of the well-known U-Net convolutional neural network. It is worth stressing that building and road classes may coexist within the same pixel when working at such a low spatial resolution, setting a challenging problem scheme. Accordingly, a robust experimental study is developed to assess the benefits of the decomposition strategies and their combination with a multi-task learning scheme. The obtained results demonstrate that decomposing the considered multi-class remote sensing semantic segmentation problem into multiple binary ones using a One-vs-All binary decomposition technique leads to better results than the standard direct multi-class approach. Additionally, the benefits of using a multi-task learning scheme for pushing the performance of binary segmentation models are also shown.
publishDate 2021
dc.date.none.fl_str_mv 2021
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv https://hdl.handle.net/2454/41735
url https://hdl.handle.net/2454/41735
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv info:eu-repo/grantAgreement/AEI/Plan Estatal de Investigación Científica y Técnica y de Innovación 2017-2020/PID2019-108392GB-I00
dc.rights.none.fl_str_mv https://creativecommons.org/licenses/by/4.0/
info:eu-repo/semantics/openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by/4.0/
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv MDPI
publisher.none.fl_str_mv MDPI
dc.source.none.fl_str_mv reponame:Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
instname:Universidad Pública de Navarra
instname_str Universidad Pública de Navarra
reponame_str Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
collection Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869417456084189184
spelling Multi-class strategies for joint building footprint and road detection in remote sensingAyala Lauroba, ChristianAranda, CarlosGalar Idoate, MikelSentinel-1Sentinel-2Remote sensingBuilding detectionRoad detectionDeep learningConvolutional neural networksMulti-class semantic segmentationBinary semantic segmentationMulti-task semantic segmentationBuilding footprints and road networks are important inputs for a great deal of services. For instance, building maps are useful for urban planning, whereas road maps are essential for disaster response services. Traditionally, building and road maps are manually generated by remote sensing experts or land surveying, occasionally assisted by semi-automatic tools. In the last decade, deep learning-based approaches have demonstrated their capabilities to extract these elements automatically and accurately from remote sensing imagery. The building footprint and road network detection problem can be considered a multi-class semantic segmentation task, that is, a single model performs a pixel-wise classification on multiple classes, optimizing the overall performance. However, depending on the spatial resolution of the imagery used, both classes may coexist within the same pixel, drastically reducing their separability. In this regard, binary decomposition techniques, which have been widely studied in the machine learning literature, are proved useful for addressing multiclass problems. Accordingly, the multi-class problem can be split into multiple binary semantic segmentation sub-problems, specializing different models for each class. Nevertheless, in these cases, an aggregation step is required to obtain the final output labels. Additionally, other novel approaches, such as multi-task learning, may come in handy to further increase the performance of the binary semantic segmentation models. Since there is no certainty as to which strategy should be carried out to accurately tackle a multi-class remote sensing semantic segmentation problem, this paper performs an in-depth study to shed light on the issue. For this purpose, open-access Sentinel-1 and Sentinel-2 imagery (at 10 m) are considered for extracting buildings and roads, making use of the well-known U-Net convolutional neural network. It is worth stressing that building and road classes may coexist within the same pixel when working at such a low spatial resolution, setting a challenging problem scheme. Accordingly, a robust experimental study is developed to assess the benefits of the decomposition strategies and their combination with a multi-task learning scheme. The obtained results demonstrate that decomposing the considered multi-class remote sensing semantic segmentation problem into multiple binary ones using a One-vs-All binary decomposition technique leads to better results than the standard direct multi-class approach. Additionally, the benefits of using a multi-task learning scheme for pushing the performance of binary segmentation models are also shown.Christian Ayala was partially supported by the Goverment of Navarra under the industrial PhD program 2020 reference 0011-1408-2020-000008. Mikel Galar was partially supported by Tracasa Instrumental S.L. under projects OTRI 2018-901-073, OTRI 2019-901-091 and OTRI 2020-901-050, and by the Spanish MICIN (PID2019-108392GB-I00 / AEI / 10.13039/501100011033).MDPIInstitute of Smart Cities - ISCGobierno de Navarra / Nafarroako Gobernua, 0011-1408-2020-0000082021info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionapplication/pdfhttps://hdl.handle.net/2454/41735reponame:Academica-e. Repositorio Institucional de la Universidad Pública de Navarrainstname:Universidad Pública de NavarraInglésinfo:eu-repo/grantAgreement/AEI/Plan Estatal de Investigación Científica y Técnica y de Innovación 2017-2020/PID2019-108392GB-I00© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.https://creativecommons.org/licenses/by/4.0/info:eu-repo/semantics/openAccessoai:academica-e.unavarra.es:2454/417352026-06-17T12:41:47Z
score 15,811543