Eintrag weiter verarbeiten
Concrete Cracks Detection Based on FCN with Dilated Convolution
Gespeichert in:
Zeitschriftentitel: | Applied Sciences |
---|---|
Personen und Körperschaften: | , , , , |
In: | Applied Sciences, 9, 2019, 13, S. 2686 |
Format: | E-Article |
Sprache: | Englisch |
veröffentlicht: |
MDPI AG
|
Schlagwörter: |
author_facet |
Zhang, Jianming Lu, Chaoquan Wang, Jin Wang, Lei Yue, Xiao-Guang Zhang, Jianming Lu, Chaoquan Wang, Jin Wang, Lei Yue, Xiao-Guang |
---|---|
author |
Zhang, Jianming Lu, Chaoquan Wang, Jin Wang, Lei Yue, Xiao-Guang |
spellingShingle |
Zhang, Jianming Lu, Chaoquan Wang, Jin Wang, Lei Yue, Xiao-Guang Applied Sciences Concrete Cracks Detection Based on FCN with Dilated Convolution Fluid Flow and Transfer Processes Computer Science Applications Process Chemistry and Technology General Engineering Instrumentation General Materials Science |
author_sort |
zhang, jianming |
spelling |
Zhang, Jianming Lu, Chaoquan Wang, Jin Wang, Lei Yue, Xiao-Guang 2076-3417 MDPI AG Fluid Flow and Transfer Processes Computer Science Applications Process Chemistry and Technology General Engineering Instrumentation General Materials Science http://dx.doi.org/10.3390/app9132686 <jats:p>In civil engineering, the stability of concrete is of great significance to safety of people’s life and property, so it is necessary to detect concrete damage effectively. In this paper, we treat crack detection on concrete surface as a semantic segmentation task that distinguishes background from crack at the pixel level. Inspired by Fully Convolutional Networks (FCN), we propose a full convolution network based on dilated convolution for concrete crack detection, which consists of an encoder and a decoder. Specifically, we first used the residual network to extract the feature maps of the input image, designed the dilated convolutions with different dilation rates to extract the feature maps of different receptive fields, and fused the extracted features from multiple branches. Then, we exploited the stacked deconvolution to do up-sampling operator in the fused feature maps. Finally, we used the SoftMax function to classify the feature maps at the pixel level. In order to verify the validity of the model, we introduced the commonly used evaluation indicators of semantic segmentation: Pixel Accuracy (PA), Mean Pixel Accuracy (MPA), Mean Intersection over Union (MIoU), and Frequency Weighted Intersection over Union (FWIoU). The experimental results show that the proposed model converges faster and has better generalization performance on the test set by introducing dilated convolutions with different dilation rates and a multi-branch fusion strategy. Our model has a PA of 96.84%, MPA of 92.55%, MIoU of 86.05% and FWIoU of 94.22% on the test set, which is superior to other models.</jats:p> Concrete Cracks Detection Based on FCN with Dilated Convolution Applied Sciences |
doi_str_mv |
10.3390/app9132686 |
facet_avail |
Online Free |
finc_class_facet |
Physik Informatik Chemie und Pharmazie Technik Allgemeines |
format |
ElectronicArticle |
fullrecord |
blob:ai-49-aHR0cDovL2R4LmRvaS5vcmcvMTAuMzM5MC9hcHA5MTMyNjg2 |
id |
ai-49-aHR0cDovL2R4LmRvaS5vcmcvMTAuMzM5MC9hcHA5MTMyNjg2 |
institution |
DE-Gla1 DE-Zi4 DE-15 DE-Pl11 DE-Rs1 DE-105 DE-14 DE-Ch1 DE-L229 DE-D275 DE-Bn3 DE-Brt1 DE-Zwi2 DE-D161 |
imprint |
MDPI AG, 2019 |
imprint_str_mv |
MDPI AG, 2019 |
issn |
2076-3417 |
issn_str_mv |
2076-3417 |
language |
English |
mega_collection |
MDPI AG (CrossRef) |
match_str |
zhang2019concretecracksdetectionbasedonfcnwithdilatedconvolution |
publishDateSort |
2019 |
publisher |
MDPI AG |
recordtype |
ai |
record_format |
ai |
series |
Applied Sciences |
source_id |
49 |
title |
Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_unstemmed |
Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_full |
Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_fullStr |
Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_full_unstemmed |
Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_short |
Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_sort |
concrete cracks detection based on fcn with dilated convolution |
topic |
Fluid Flow and Transfer Processes Computer Science Applications Process Chemistry and Technology General Engineering Instrumentation General Materials Science |
url |
http://dx.doi.org/10.3390/app9132686 |
publishDate |
2019 |
physical |
2686 |
description |
<jats:p>In civil engineering, the stability of concrete is of great significance to safety of people’s life and property, so it is necessary to detect concrete damage effectively. In this paper, we treat crack detection on concrete surface as a semantic segmentation task that distinguishes background from crack at the pixel level. Inspired by Fully Convolutional Networks (FCN), we propose a full convolution network based on dilated convolution for concrete crack detection, which consists of an encoder and a decoder. Specifically, we first used the residual network to extract the feature maps of the input image, designed the dilated convolutions with different dilation rates to extract the feature maps of different receptive fields, and fused the extracted features from multiple branches. Then, we exploited the stacked deconvolution to do up-sampling operator in the fused feature maps. Finally, we used the SoftMax function to classify the feature maps at the pixel level. In order to verify the validity of the model, we introduced the commonly used evaluation indicators of semantic segmentation: Pixel Accuracy (PA), Mean Pixel Accuracy (MPA), Mean Intersection over Union (MIoU), and Frequency Weighted Intersection over Union (FWIoU). The experimental results show that the proposed model converges faster and has better generalization performance on the test set by introducing dilated convolutions with different dilation rates and a multi-branch fusion strategy. Our model has a PA of 96.84%, MPA of 92.55%, MIoU of 86.05% and FWIoU of 94.22% on the test set, which is superior to other models.</jats:p> |
container_issue |
13 |
container_start_page |
0 |
container_title |
Applied Sciences |
container_volume |
9 |
format_de105 |
Article, E-Article |
format_de14 |
Article, E-Article |
format_de15 |
Article, E-Article |
format_de520 |
Article, E-Article |
format_de540 |
Article, E-Article |
format_dech1 |
Article, E-Article |
format_ded117 |
Article, E-Article |
format_degla1 |
E-Article |
format_del152 |
Buch |
format_del189 |
Article, E-Article |
format_dezi4 |
Article |
format_dezwi2 |
Article, E-Article |
format_finc |
Article, E-Article |
format_nrw |
Article, E-Article |
_version_ |
1792347067372273674 |
geogr_code |
not assigned |
last_indexed |
2024-03-01T17:49:24.209Z |
geogr_code_person |
not assigned |
openURL |
url_ver=Z39.88-2004&ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fvufind.svn.sourceforge.net%3Agenerator&rft.title=Concrete+Cracks+Detection+Based+on+FCN+with+Dilated+Convolution&rft.date=2019-07-01&genre=article&issn=2076-3417&volume=9&issue=13&pages=2686&jtitle=Applied+Sciences&atitle=Concrete+Cracks+Detection+Based+on+FCN+with+Dilated+Convolution&aulast=Yue&aufirst=Xiao-Guang&rft_id=info%3Adoi%2F10.3390%2Fapp9132686&rft.language%5B0%5D=eng |
SOLR | |
_version_ | 1792347067372273674 |
author | Zhang, Jianming, Lu, Chaoquan, Wang, Jin, Wang, Lei, Yue, Xiao-Guang |
author_facet | Zhang, Jianming, Lu, Chaoquan, Wang, Jin, Wang, Lei, Yue, Xiao-Guang, Zhang, Jianming, Lu, Chaoquan, Wang, Jin, Wang, Lei, Yue, Xiao-Guang |
author_sort | zhang, jianming |
container_issue | 13 |
container_start_page | 0 |
container_title | Applied Sciences |
container_volume | 9 |
description | <jats:p>In civil engineering, the stability of concrete is of great significance to safety of people’s life and property, so it is necessary to detect concrete damage effectively. In this paper, we treat crack detection on concrete surface as a semantic segmentation task that distinguishes background from crack at the pixel level. Inspired by Fully Convolutional Networks (FCN), we propose a full convolution network based on dilated convolution for concrete crack detection, which consists of an encoder and a decoder. Specifically, we first used the residual network to extract the feature maps of the input image, designed the dilated convolutions with different dilation rates to extract the feature maps of different receptive fields, and fused the extracted features from multiple branches. Then, we exploited the stacked deconvolution to do up-sampling operator in the fused feature maps. Finally, we used the SoftMax function to classify the feature maps at the pixel level. In order to verify the validity of the model, we introduced the commonly used evaluation indicators of semantic segmentation: Pixel Accuracy (PA), Mean Pixel Accuracy (MPA), Mean Intersection over Union (MIoU), and Frequency Weighted Intersection over Union (FWIoU). The experimental results show that the proposed model converges faster and has better generalization performance on the test set by introducing dilated convolutions with different dilation rates and a multi-branch fusion strategy. Our model has a PA of 96.84%, MPA of 92.55%, MIoU of 86.05% and FWIoU of 94.22% on the test set, which is superior to other models.</jats:p> |
doi_str_mv | 10.3390/app9132686 |
facet_avail | Online, Free |
finc_class_facet | Physik, Informatik, Chemie und Pharmazie, Technik, Allgemeines |
format | ElectronicArticle |
format_de105 | Article, E-Article |
format_de14 | Article, E-Article |
format_de15 | Article, E-Article |
format_de520 | Article, E-Article |
format_de540 | Article, E-Article |
format_dech1 | Article, E-Article |
format_ded117 | Article, E-Article |
format_degla1 | E-Article |
format_del152 | Buch |
format_del189 | Article, E-Article |
format_dezi4 | Article |
format_dezwi2 | Article, E-Article |
format_finc | Article, E-Article |
format_nrw | Article, E-Article |
geogr_code | not assigned |
geogr_code_person | not assigned |
id | ai-49-aHR0cDovL2R4LmRvaS5vcmcvMTAuMzM5MC9hcHA5MTMyNjg2 |
imprint | MDPI AG, 2019 |
imprint_str_mv | MDPI AG, 2019 |
institution | DE-Gla1, DE-Zi4, DE-15, DE-Pl11, DE-Rs1, DE-105, DE-14, DE-Ch1, DE-L229, DE-D275, DE-Bn3, DE-Brt1, DE-Zwi2, DE-D161 |
issn | 2076-3417 |
issn_str_mv | 2076-3417 |
language | English |
last_indexed | 2024-03-01T17:49:24.209Z |
match_str | zhang2019concretecracksdetectionbasedonfcnwithdilatedconvolution |
mega_collection | MDPI AG (CrossRef) |
physical | 2686 |
publishDate | 2019 |
publishDateSort | 2019 |
publisher | MDPI AG |
record_format | ai |
recordtype | ai |
series | Applied Sciences |
source_id | 49 |
spelling | Zhang, Jianming Lu, Chaoquan Wang, Jin Wang, Lei Yue, Xiao-Guang 2076-3417 MDPI AG Fluid Flow and Transfer Processes Computer Science Applications Process Chemistry and Technology General Engineering Instrumentation General Materials Science http://dx.doi.org/10.3390/app9132686 <jats:p>In civil engineering, the stability of concrete is of great significance to safety of people’s life and property, so it is necessary to detect concrete damage effectively. In this paper, we treat crack detection on concrete surface as a semantic segmentation task that distinguishes background from crack at the pixel level. Inspired by Fully Convolutional Networks (FCN), we propose a full convolution network based on dilated convolution for concrete crack detection, which consists of an encoder and a decoder. Specifically, we first used the residual network to extract the feature maps of the input image, designed the dilated convolutions with different dilation rates to extract the feature maps of different receptive fields, and fused the extracted features from multiple branches. Then, we exploited the stacked deconvolution to do up-sampling operator in the fused feature maps. Finally, we used the SoftMax function to classify the feature maps at the pixel level. In order to verify the validity of the model, we introduced the commonly used evaluation indicators of semantic segmentation: Pixel Accuracy (PA), Mean Pixel Accuracy (MPA), Mean Intersection over Union (MIoU), and Frequency Weighted Intersection over Union (FWIoU). The experimental results show that the proposed model converges faster and has better generalization performance on the test set by introducing dilated convolutions with different dilation rates and a multi-branch fusion strategy. Our model has a PA of 96.84%, MPA of 92.55%, MIoU of 86.05% and FWIoU of 94.22% on the test set, which is superior to other models.</jats:p> Concrete Cracks Detection Based on FCN with Dilated Convolution Applied Sciences |
spellingShingle | Zhang, Jianming, Lu, Chaoquan, Wang, Jin, Wang, Lei, Yue, Xiao-Guang, Applied Sciences, Concrete Cracks Detection Based on FCN with Dilated Convolution, Fluid Flow and Transfer Processes, Computer Science Applications, Process Chemistry and Technology, General Engineering, Instrumentation, General Materials Science |
title | Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_full | Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_fullStr | Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_full_unstemmed | Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_short | Concrete Cracks Detection Based on FCN with Dilated Convolution |
title_sort | concrete cracks detection based on fcn with dilated convolution |
title_unstemmed | Concrete Cracks Detection Based on FCN with Dilated Convolution |
topic | Fluid Flow and Transfer Processes, Computer Science Applications, Process Chemistry and Technology, General Engineering, Instrumentation, General Materials Science |
url | http://dx.doi.org/10.3390/app9132686 |