A Novel Siamese-Based Approach for Scene Change Detection With Applications to Obstructed Routes in Hazardous Environments
Autor(a) principal: | |
---|---|
Data de Publicação: | 2020 |
Outros Autores: | , , , , |
Tipo de documento: | Artigo |
Idioma: | eng |
Título da fonte: | Repositório Institucional da UNESP |
Texto Completo: | http://dx.doi.org/10.1109/MIS.2019.2949984 http://hdl.handle.net/11449/197720 |
Resumo: | The demand for automatic scene change detection has massively increased in the last decades due to its importance regarding safety and security issues. Although deep learning techniques have provided significant enhancements in the field, such methods must learn which object belongs to the foreground or background beforehand. In this article, we propose an approach that employs siamese U-Nets to address the task of change detection, such that the model learns to perform semantic segmentation using background reference frames only. Therefore, any object that comes up into the scene defines a change. The experimental results show the robustness of the proposed model over the well-known public dataset CDNet2014. Additionally, we also consider a private dataset called PetrobrasROUTES, which comprises obstruction or abandoned objects in escape routes in hazardous environments. Moreover, the experiments show that the proposed approach is more robust to noise and illumination changes. |
id |
UNSP_46fb92627c05687d44bd375356cdc061 |
---|---|
oai_identifier_str |
oai:repositorio.unesp.br:11449/197720 |
network_acronym_str |
UNSP |
network_name_str |
Repositório Institucional da UNESP |
repository_id_str |
2946 |
spelling |
A Novel Siamese-Based Approach for Scene Change Detection With Applications to Obstructed Routes in Hazardous EnvironmentsDecodingImage segmentationSemanticsTraining dataNeural networksIntelligent systemsTask analysisHuman computer interactionScene Change DetectionSiamese Convolutional Neural NetworksU-NetsRoute Obstruction DetectionThe demand for automatic scene change detection has massively increased in the last decades due to its importance regarding safety and security issues. Although deep learning techniques have provided significant enhancements in the field, such methods must learn which object belongs to the foreground or background beforehand. In this article, we propose an approach that employs siamese U-Nets to address the task of change detection, such that the model learns to perform semantic segmentation using background reference frames only. Therefore, any object that comes up into the scene defines a change. The experimental results show the robustness of the proposed model over the well-known public dataset CDNet2014. Additionally, we also consider a private dataset called PetrobrasROUTES, which comprises obstruction or abandoned objects in escape routes in hazardous environments. Moreover, the experiments show that the proposed approach is more robust to noise and illumination changes.Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)PetrobrasSao Paulo State Univ, Sao Paulo, BrazilPetr Brasileiro SA Petrobras, Rio De Janeiro, BrazilUniv Fortaleza, UNIFOR, Fortaleza, Ceara, BrazilSao Paulo State Univ, Sao Paulo, BrazilFAPESP: 2013/07375-0FAPESP: 2014/12236-1FAPESP: 2016/19403-6FAPESP: 2017/25908-6CNPq: 307066/2017-7CNPq: 427968/2018-6CNPq: 304315/2017-6CNPq: 430274/2018-1Petrobras: 2017/00285-6Ieee Computer SocUniversidade Estadual Paulista (Unesp)Petr Brasileiro SA PetrobrasUniv FortalezaSantana, Marcos C. S. [UNESP]Passos, Leandro Aparecido [UNESP]Moreira, Thierry P. [UNESP]Colombo, DaniloAlbuquerque, Victor Hugo C. dePapa, Joao Paulo [UNESP]2020-12-11T14:35:48Z2020-12-11T14:35:48Z2020-01-01info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/article44-53http://dx.doi.org/10.1109/MIS.2019.2949984Ieee Intelligent Systems. Los Alamitos: Ieee Computer Soc, v. 35, n. 1, p. 44-53, 2020.1541-1672http://hdl.handle.net/11449/19772010.1109/MIS.2019.2949984WOS:000522198900006Web of Sciencereponame:Repositório Institucional da UNESPinstname:Universidade Estadual Paulista (UNESP)instacron:UNESPengIeee Intelligent Systemsinfo:eu-repo/semantics/openAccess2024-04-23T16:10:45Zoai:repositorio.unesp.br:11449/197720Repositório InstitucionalPUBhttp://repositorio.unesp.br/oai/requestopendoar:29462024-08-05T16:17:40.602836Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)false |
dc.title.none.fl_str_mv |
A Novel Siamese-Based Approach for Scene Change Detection With Applications to Obstructed Routes in Hazardous Environments |
title |
A Novel Siamese-Based Approach for Scene Change Detection With Applications to Obstructed Routes in Hazardous Environments |
spellingShingle |
A Novel Siamese-Based Approach for Scene Change Detection With Applications to Obstructed Routes in Hazardous Environments Santana, Marcos C. S. [UNESP] Decoding Image segmentation Semantics Training data Neural networks Intelligent systems Task analysis Human computer interaction Scene Change Detection Siamese Convolutional Neural Networks U-Nets Route Obstruction Detection |
title_short |
A Novel Siamese-Based Approach for Scene Change Detection With Applications to Obstructed Routes in Hazardous Environments |
title_full |
A Novel Siamese-Based Approach for Scene Change Detection With Applications to Obstructed Routes in Hazardous Environments |
title_fullStr |
A Novel Siamese-Based Approach for Scene Change Detection With Applications to Obstructed Routes in Hazardous Environments |
title_full_unstemmed |
A Novel Siamese-Based Approach for Scene Change Detection With Applications to Obstructed Routes in Hazardous Environments |
title_sort |
A Novel Siamese-Based Approach for Scene Change Detection With Applications to Obstructed Routes in Hazardous Environments |
author |
Santana, Marcos C. S. [UNESP] |
author_facet |
Santana, Marcos C. S. [UNESP] Passos, Leandro Aparecido [UNESP] Moreira, Thierry P. [UNESP] Colombo, Danilo Albuquerque, Victor Hugo C. de Papa, Joao Paulo [UNESP] |
author_role |
author |
author2 |
Passos, Leandro Aparecido [UNESP] Moreira, Thierry P. [UNESP] Colombo, Danilo Albuquerque, Victor Hugo C. de Papa, Joao Paulo [UNESP] |
author2_role |
author author author author author |
dc.contributor.none.fl_str_mv |
Universidade Estadual Paulista (Unesp) Petr Brasileiro SA Petrobras Univ Fortaleza |
dc.contributor.author.fl_str_mv |
Santana, Marcos C. S. [UNESP] Passos, Leandro Aparecido [UNESP] Moreira, Thierry P. [UNESP] Colombo, Danilo Albuquerque, Victor Hugo C. de Papa, Joao Paulo [UNESP] |
dc.subject.por.fl_str_mv |
Decoding Image segmentation Semantics Training data Neural networks Intelligent systems Task analysis Human computer interaction Scene Change Detection Siamese Convolutional Neural Networks U-Nets Route Obstruction Detection |
topic |
Decoding Image segmentation Semantics Training data Neural networks Intelligent systems Task analysis Human computer interaction Scene Change Detection Siamese Convolutional Neural Networks U-Nets Route Obstruction Detection |
description |
The demand for automatic scene change detection has massively increased in the last decades due to its importance regarding safety and security issues. Although deep learning techniques have provided significant enhancements in the field, such methods must learn which object belongs to the foreground or background beforehand. In this article, we propose an approach that employs siamese U-Nets to address the task of change detection, such that the model learns to perform semantic segmentation using background reference frames only. Therefore, any object that comes up into the scene defines a change. The experimental results show the robustness of the proposed model over the well-known public dataset CDNet2014. Additionally, we also consider a private dataset called PetrobrasROUTES, which comprises obstruction or abandoned objects in escape routes in hazardous environments. Moreover, the experiments show that the proposed approach is more robust to noise and illumination changes. |
publishDate |
2020 |
dc.date.none.fl_str_mv |
2020-12-11T14:35:48Z 2020-12-11T14:35:48Z 2020-01-01 |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/article |
format |
article |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://dx.doi.org/10.1109/MIS.2019.2949984 Ieee Intelligent Systems. Los Alamitos: Ieee Computer Soc, v. 35, n. 1, p. 44-53, 2020. 1541-1672 http://hdl.handle.net/11449/197720 10.1109/MIS.2019.2949984 WOS:000522198900006 |
url |
http://dx.doi.org/10.1109/MIS.2019.2949984 http://hdl.handle.net/11449/197720 |
identifier_str_mv |
Ieee Intelligent Systems. Los Alamitos: Ieee Computer Soc, v. 35, n. 1, p. 44-53, 2020. 1541-1672 10.1109/MIS.2019.2949984 WOS:000522198900006 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
Ieee Intelligent Systems |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
44-53 |
dc.publisher.none.fl_str_mv |
Ieee Computer Soc |
publisher.none.fl_str_mv |
Ieee Computer Soc |
dc.source.none.fl_str_mv |
Web of Science reponame:Repositório Institucional da UNESP instname:Universidade Estadual Paulista (UNESP) instacron:UNESP |
instname_str |
Universidade Estadual Paulista (UNESP) |
instacron_str |
UNESP |
institution |
UNESP |
reponame_str |
Repositório Institucional da UNESP |
collection |
Repositório Institucional da UNESP |
repository.name.fl_str_mv |
Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP) |
repository.mail.fl_str_mv |
|
_version_ |
1808128629675130880 |