Cargando…

Improved Pixel-Level Pavement-Defect Segmentation Using a Deep Autoencoder

Convolutional neural networks perform impressively in complicated computer-vision image-segmentation tasks. Vision-based systems surpass humans in speed and accuracy in quality inspection tasks. Moreover, the maintenance of big infrastructures, such as roads, bridges, or buildings, is tedious and ti...

Descripción completa

Detalles Bibliográficos
Autores principales: Augustauskas, Rytis, Lipnickas, Arūnas
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2020
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7249178/
https://www.ncbi.nlm.nih.gov/pubmed/32365925
http://dx.doi.org/10.3390/s20092557
_version_ 1783538544265920512
author Augustauskas, Rytis
Lipnickas, Arūnas
author_facet Augustauskas, Rytis
Lipnickas, Arūnas
author_sort Augustauskas, Rytis
collection PubMed
description Convolutional neural networks perform impressively in complicated computer-vision image-segmentation tasks. Vision-based systems surpass humans in speed and accuracy in quality inspection tasks. Moreover, the maintenance of big infrastructures, such as roads, bridges, or buildings, is tedious and time-demanding work. In this research, we addressed pavement-quality evaluation by pixelwise defect segmentation using a U-Net deep autoencoder. Additionally, to the original neural network architecture, we utilized residual connections, atrous spatial pyramid pooling with parallel and “Waterfall” connections, and attention gates to perform better defect extraction. The proposed neural network configurations showed a segmentation performance improvement over U-Net with no significant computational overhead. Statistical and visual performance evaluation was taken into consideration for the model comparison. Experiments were conducted on CrackForest, Crack500, GAPs384, and mixed datasets.
format Online
Article
Text
id pubmed-7249178
institution National Center for Biotechnology Information
language English
publishDate 2020
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-72491782020-06-10 Improved Pixel-Level Pavement-Defect Segmentation Using a Deep Autoencoder Augustauskas, Rytis Lipnickas, Arūnas Sensors (Basel) Article Convolutional neural networks perform impressively in complicated computer-vision image-segmentation tasks. Vision-based systems surpass humans in speed and accuracy in quality inspection tasks. Moreover, the maintenance of big infrastructures, such as roads, bridges, or buildings, is tedious and time-demanding work. In this research, we addressed pavement-quality evaluation by pixelwise defect segmentation using a U-Net deep autoencoder. Additionally, to the original neural network architecture, we utilized residual connections, atrous spatial pyramid pooling with parallel and “Waterfall” connections, and attention gates to perform better defect extraction. The proposed neural network configurations showed a segmentation performance improvement over U-Net with no significant computational overhead. Statistical and visual performance evaluation was taken into consideration for the model comparison. Experiments were conducted on CrackForest, Crack500, GAPs384, and mixed datasets. MDPI 2020-04-30 /pmc/articles/PMC7249178/ /pubmed/32365925 http://dx.doi.org/10.3390/s20092557 Text en © 2020 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 (http://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Augustauskas, Rytis
Lipnickas, Arūnas
Improved Pixel-Level Pavement-Defect Segmentation Using a Deep Autoencoder
title Improved Pixel-Level Pavement-Defect Segmentation Using a Deep Autoencoder
title_full Improved Pixel-Level Pavement-Defect Segmentation Using a Deep Autoencoder
title_fullStr Improved Pixel-Level Pavement-Defect Segmentation Using a Deep Autoencoder
title_full_unstemmed Improved Pixel-Level Pavement-Defect Segmentation Using a Deep Autoencoder
title_short Improved Pixel-Level Pavement-Defect Segmentation Using a Deep Autoencoder
title_sort improved pixel-level pavement-defect segmentation using a deep autoencoder
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7249178/
https://www.ncbi.nlm.nih.gov/pubmed/32365925
http://dx.doi.org/10.3390/s20092557
work_keys_str_mv AT augustauskasrytis improvedpixellevelpavementdefectsegmentationusingadeepautoencoder
AT lipnickasarunas improvedpixellevelpavementdefectsegmentationusingadeepautoencoder