Automated Ground Truth Generation for Learning-Based Crack Detection on Concrete Surfaces

This article introduces an automated data-labeling approach for generating crack ground truths (GTs) within concrete images. The main algorithm includes generating first-round GTs, pre-training a deep learning-based model, and generating second-round GTs. On the basis of the generated second-round G...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Hsiang-Chieh Chen, Zheng-Ting Li
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	automated data labeling crack detection crack segmentation deep learning ground truth generation Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
Acceso en línea:	https://doaj.org/article/ebcf0a49da7240aa9738c10a4bfdbc2c
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:ebcf0a49da7240aa9738c10a4bfdbc2c
record_format	dspace
spelling	oai:doaj.org-article:ebcf0a49da7240aa9738c10a4bfdbc2c2021-11-25T16:42:01ZAutomated Ground Truth Generation for Learning-Based Crack Detection on Concrete Surfaces10.3390/app1122109662076-3417https://doaj.org/article/ebcf0a49da7240aa9738c10a4bfdbc2c2021-11-01T00:00:00Zhttps://www.mdpi.com/2076-3417/11/22/10966https://doaj.org/toc/2076-3417This article introduces an automated data-labeling approach for generating crack ground truths (GTs) within concrete images. The main algorithm includes generating first-round GTs, pre-training a deep learning-based model, and generating second-round GTs. On the basis of the generated second-round GTs of the training data, a learning-based crack detection model can be trained in a self-supervised manner. The pre-trained deep learning-based model is effective for crack detection after it is re-trained using the second-round GTs. The main contribution of this study is the proposal of an automated GT generation process for training a crack detection model at the pixel level. Experimental results show that the second-round GTs are similar to manually marked labels. Accordingly, the cost of implementing learning-based methods is reduced significantly because data labeling by humans is not necessitated.Hsiang-Chieh ChenZheng-Ting LiMDPI AGarticleautomated data labelingcrack detectioncrack segmentationdeep learningground truth generationTechnologyTEngineering (General). Civil engineering (General)TA1-2040Biology (General)QH301-705.5PhysicsQC1-999ChemistryQD1-999ENApplied Sciences, Vol 11, Iss 10966, p 10966 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	automated data labeling crack detection crack segmentation deep learning ground truth generation Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
spellingShingle	automated data labeling crack detection crack segmentation deep learning ground truth generation Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999 Hsiang-Chieh Chen Zheng-Ting Li Automated Ground Truth Generation for Learning-Based Crack Detection on Concrete Surfaces
description	This article introduces an automated data-labeling approach for generating crack ground truths (GTs) within concrete images. The main algorithm includes generating first-round GTs, pre-training a deep learning-based model, and generating second-round GTs. On the basis of the generated second-round GTs of the training data, a learning-based crack detection model can be trained in a self-supervised manner. The pre-trained deep learning-based model is effective for crack detection after it is re-trained using the second-round GTs. The main contribution of this study is the proposal of an automated GT generation process for training a crack detection model at the pixel level. Experimental results show that the second-round GTs are similar to manually marked labels. Accordingly, the cost of implementing learning-based methods is reduced significantly because data labeling by humans is not necessitated.
format	article
author	Hsiang-Chieh Chen Zheng-Ting Li
author_facet	Hsiang-Chieh Chen Zheng-Ting Li
author_sort	Hsiang-Chieh Chen
title	Automated Ground Truth Generation for Learning-Based Crack Detection on Concrete Surfaces
title_short	Automated Ground Truth Generation for Learning-Based Crack Detection on Concrete Surfaces
title_full	Automated Ground Truth Generation for Learning-Based Crack Detection on Concrete Surfaces
title_fullStr	Automated Ground Truth Generation for Learning-Based Crack Detection on Concrete Surfaces
title_full_unstemmed	Automated Ground Truth Generation for Learning-Based Crack Detection on Concrete Surfaces
title_sort	automated ground truth generation for learning-based crack detection on concrete surfaces
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/ebcf0a49da7240aa9738c10a4bfdbc2c
work_keys_str_mv	AT hsiangchiehchen automatedgroundtruthgenerationforlearningbasedcrackdetectiononconcretesurfaces AT zhengtingli automatedgroundtruthgenerationforlearningbasedcrackdetectiononconcretesurfaces
_version_	1718413052275064832

Automated Ground Truth Generation for Learning-Based Crack Detection on Concrete Surfaces

Ejemplares similares