Cargando…

Wall segmentation in 2D images using convolutional neural networks

Wall segmentation is a special case of semantic segmentation, and the task is to classify each pixel into one of two classes: wall and no-wall. The segmentation model returns a mask showing where objects like windows and furniture are located, as well as walls. This article proposes the module’s str...

Descripción completa

Detalles Bibliográficos
Autores principales: Bjekic, Mihailo, Lazovic, Ana, K, Venkatachalam, Bacanin, Nebojsa, Zivkovic, Miodrag, Kvascev, Goran, Nikolic, Bosko
Formato: Online Artículo Texto
Lenguaje:English
Publicado: PeerJ Inc. 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10557507/
https://www.ncbi.nlm.nih.gov/pubmed/37810356
http://dx.doi.org/10.7717/peerj-cs.1565
_version_ 1785117104708517888
author Bjekic, Mihailo
Lazovic, Ana
K, Venkatachalam
Bacanin, Nebojsa
Zivkovic, Miodrag
Kvascev, Goran
Nikolic, Bosko
author_facet Bjekic, Mihailo
Lazovic, Ana
K, Venkatachalam
Bacanin, Nebojsa
Zivkovic, Miodrag
Kvascev, Goran
Nikolic, Bosko
author_sort Bjekic, Mihailo
collection PubMed
description Wall segmentation is a special case of semantic segmentation, and the task is to classify each pixel into one of two classes: wall and no-wall. The segmentation model returns a mask showing where objects like windows and furniture are located, as well as walls. This article proposes the module’s structure for semantic segmentation of walls in 2D images, which can effectively address the problem of wall segmentation. The proposed model achieved higher accuracy and faster execution than other solutions. An encoder-decoder architecture of the segmentation module was used. Dilated ResNet50/101 network was used as an encoder, representing ResNet50/101 network in which dilated convolutional layers replaced the last convolutional layers. The ADE20K dataset subset containing only interior images, was used for model training, while only its subset was used for model evaluation. Three different approaches to model training were analyzed in the research. On the validation dataset, the best approach based on the proposed structure with the ResNet101 network resulted in an average accuracy at the pixel level of 92.13% and an intersection over union (IoU) of 72.58%. Moreover, all proposed approaches can be applied to recognize other objects in the image to solve specific tasks.
format Online
Article
Text
id pubmed-10557507
institution National Center for Biotechnology Information
language English
publishDate 2023
publisher PeerJ Inc.
record_format MEDLINE/PubMed
spelling pubmed-105575072023-10-07 Wall segmentation in 2D images using convolutional neural networks Bjekic, Mihailo Lazovic, Ana K, Venkatachalam Bacanin, Nebojsa Zivkovic, Miodrag Kvascev, Goran Nikolic, Bosko PeerJ Comput Sci Artificial Intelligence Wall segmentation is a special case of semantic segmentation, and the task is to classify each pixel into one of two classes: wall and no-wall. The segmentation model returns a mask showing where objects like windows and furniture are located, as well as walls. This article proposes the module’s structure for semantic segmentation of walls in 2D images, which can effectively address the problem of wall segmentation. The proposed model achieved higher accuracy and faster execution than other solutions. An encoder-decoder architecture of the segmentation module was used. Dilated ResNet50/101 network was used as an encoder, representing ResNet50/101 network in which dilated convolutional layers replaced the last convolutional layers. The ADE20K dataset subset containing only interior images, was used for model training, while only its subset was used for model evaluation. Three different approaches to model training were analyzed in the research. On the validation dataset, the best approach based on the proposed structure with the ResNet101 network resulted in an average accuracy at the pixel level of 92.13% and an intersection over union (IoU) of 72.58%. Moreover, all proposed approaches can be applied to recognize other objects in the image to solve specific tasks. PeerJ Inc. 2023-09-11 /pmc/articles/PMC10557507/ /pubmed/37810356 http://dx.doi.org/10.7717/peerj-cs.1565 Text en © 2023 Bjekic et al. https://creativecommons.org/licenses/by/4.0/This is an open access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/) , which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. For attribution, the original author(s), title, publication source (PeerJ Computer Science) and either DOI or URL of the article must be cited.
spellingShingle Artificial Intelligence
Bjekic, Mihailo
Lazovic, Ana
K, Venkatachalam
Bacanin, Nebojsa
Zivkovic, Miodrag
Kvascev, Goran
Nikolic, Bosko
Wall segmentation in 2D images using convolutional neural networks
title Wall segmentation in 2D images using convolutional neural networks
title_full Wall segmentation in 2D images using convolutional neural networks
title_fullStr Wall segmentation in 2D images using convolutional neural networks
title_full_unstemmed Wall segmentation in 2D images using convolutional neural networks
title_short Wall segmentation in 2D images using convolutional neural networks
title_sort wall segmentation in 2d images using convolutional neural networks
topic Artificial Intelligence
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10557507/
https://www.ncbi.nlm.nih.gov/pubmed/37810356
http://dx.doi.org/10.7717/peerj-cs.1565
work_keys_str_mv AT bjekicmihailo wallsegmentationin2dimagesusingconvolutionalneuralnetworks
AT lazovicana wallsegmentationin2dimagesusingconvolutionalneuralnetworks
AT kvenkatachalam wallsegmentationin2dimagesusingconvolutionalneuralnetworks
AT bacaninnebojsa wallsegmentationin2dimagesusingconvolutionalneuralnetworks
AT zivkovicmiodrag wallsegmentationin2dimagesusingconvolutionalneuralnetworks
AT kvascevgoran wallsegmentationin2dimagesusingconvolutionalneuralnetworks
AT nikolicbosko wallsegmentationin2dimagesusingconvolutionalneuralnetworks