A hybrid attention deep learning network for refined segmentation of cracks from shield tunnel lining images
This research developed a hybrid position-channel network (named PCNet) through incorporating newly designed channel and position attention modules into U-Net to alleviate the crack discontinuity problem in channel and spatial dimensions. In PCNet, the U-Net is used as a baseline to extract informat...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-12-01
|
| Series: | Journal of Rock Mechanics and Geotechnical Engineering |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1674775523001117 |
| _version_ | 1827616171440996352 |
|---|---|
| author | Shuai Zhao Guokai Zhang Dongming Zhang Daoyuan Tan Hongwei Huang |
| author_facet | Shuai Zhao Guokai Zhang Dongming Zhang Daoyuan Tan Hongwei Huang |
| author_sort | Shuai Zhao |
| collection | DOAJ |
| description | This research developed a hybrid position-channel network (named PCNet) through incorporating newly designed channel and position attention modules into U-Net to alleviate the crack discontinuity problem in channel and spatial dimensions. In PCNet, the U-Net is used as a baseline to extract informative spatial and channel-wise features from shield tunnel lining crack images. A channel and a position attention module are designed and embedded after each convolution layer of U-Net to model the feature interdependencies in channel and spatial dimensions. These attention modules can make the U-Net adaptively integrate local crack features with their global dependencies. Experiments were conducted utilizing the dataset based on the images from Shanghai metro shield tunnels. The results validate the effectiveness of the designed channel and position attention modules, since they can individually increase balanced accuracy (BA) by 11.25% and 12.95%, intersection over union (IoU) by 10.79% and 11.83%, and F1 score by 9.96% and 10.63%, respectively. In comparison with the state-of-the-art models (i.e. LinkNet, PSPNet, U-Net, PANet, and Mask R–CNN) on the testing dataset, the proposed PCNet outperforms others with an improvement of BA, IoU, and F1 score owing to the implementation of the channel and position attention modules. These evaluation metrics indicate that the proposed PCNet presents refined crack segmentation with improved performance and is a practicable approach to segment shield tunnel lining cracks in field practice. |
| first_indexed | 2024-03-09T09:22:21Z |
| format | Article |
| id | doaj.art-a3e6a282b7fe4ba2a8f5fee819ee4212 |
| institution | Directory Open Access Journal |
| issn | 1674-7755 |
| language | English |
| last_indexed | 2024-03-09T09:22:21Z |
| publishDate | 2023-12-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Journal of Rock Mechanics and Geotechnical Engineering |
| spelling | doaj.art-a3e6a282b7fe4ba2a8f5fee819ee42122023-12-02T06:58:56ZengElsevierJournal of Rock Mechanics and Geotechnical Engineering1674-77552023-12-01151231053117A hybrid attention deep learning network for refined segmentation of cracks from shield tunnel lining imagesShuai Zhao0Guokai Zhang1Dongming Zhang2Daoyuan Tan3Hongwei Huang4Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, 999077, Hong Kong, ChinaSchool of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai, 200093, ChinaKey Laboratory of Geotechnical and Underground Engineering of Ministry of Education, Tongji University, Shanghai, 200092, ChinaDepartment of Civil and Environmental Engineering, The Hong Kong Polytechnic University, 999077, Hong Kong, China; Corresponding author.Key Laboratory of Geotechnical and Underground Engineering of Ministry of Education, Tongji University, Shanghai, 200092, ChinaThis research developed a hybrid position-channel network (named PCNet) through incorporating newly designed channel and position attention modules into U-Net to alleviate the crack discontinuity problem in channel and spatial dimensions. In PCNet, the U-Net is used as a baseline to extract informative spatial and channel-wise features from shield tunnel lining crack images. A channel and a position attention module are designed and embedded after each convolution layer of U-Net to model the feature interdependencies in channel and spatial dimensions. These attention modules can make the U-Net adaptively integrate local crack features with their global dependencies. Experiments were conducted utilizing the dataset based on the images from Shanghai metro shield tunnels. The results validate the effectiveness of the designed channel and position attention modules, since they can individually increase balanced accuracy (BA) by 11.25% and 12.95%, intersection over union (IoU) by 10.79% and 11.83%, and F1 score by 9.96% and 10.63%, respectively. In comparison with the state-of-the-art models (i.e. LinkNet, PSPNet, U-Net, PANet, and Mask R–CNN) on the testing dataset, the proposed PCNet outperforms others with an improvement of BA, IoU, and F1 score owing to the implementation of the channel and position attention modules. These evaluation metrics indicate that the proposed PCNet presents refined crack segmentation with improved performance and is a practicable approach to segment shield tunnel lining cracks in field practice.http://www.sciencedirect.com/science/article/pii/S1674775523001117Crack segmentationCrack disjoint problemU-netChannel attentionPosition attention |
| spellingShingle | Shuai Zhao Guokai Zhang Dongming Zhang Daoyuan Tan Hongwei Huang A hybrid attention deep learning network for refined segmentation of cracks from shield tunnel lining images Journal of Rock Mechanics and Geotechnical Engineering Crack segmentation Crack disjoint problem U-net Channel attention Position attention |
| title | A hybrid attention deep learning network for refined segmentation of cracks from shield tunnel lining images |
| title_full | A hybrid attention deep learning network for refined segmentation of cracks from shield tunnel lining images |
| title_fullStr | A hybrid attention deep learning network for refined segmentation of cracks from shield tunnel lining images |
| title_full_unstemmed | A hybrid attention deep learning network for refined segmentation of cracks from shield tunnel lining images |
| title_short | A hybrid attention deep learning network for refined segmentation of cracks from shield tunnel lining images |
| title_sort | hybrid attention deep learning network for refined segmentation of cracks from shield tunnel lining images |
| topic | Crack segmentation Crack disjoint problem U-net Channel attention Position attention |
| url | http://www.sciencedirect.com/science/article/pii/S1674775523001117 |
| work_keys_str_mv | AT shuaizhao ahybridattentiondeeplearningnetworkforrefinedsegmentationofcracksfromshieldtunnelliningimages AT guokaizhang ahybridattentiondeeplearningnetworkforrefinedsegmentationofcracksfromshieldtunnelliningimages AT dongmingzhang ahybridattentiondeeplearningnetworkforrefinedsegmentationofcracksfromshieldtunnelliningimages AT daoyuantan ahybridattentiondeeplearningnetworkforrefinedsegmentationofcracksfromshieldtunnelliningimages AT hongweihuang ahybridattentiondeeplearningnetworkforrefinedsegmentationofcracksfromshieldtunnelliningimages AT shuaizhao hybridattentiondeeplearningnetworkforrefinedsegmentationofcracksfromshieldtunnelliningimages AT guokaizhang hybridattentiondeeplearningnetworkforrefinedsegmentationofcracksfromshieldtunnelliningimages AT dongmingzhang hybridattentiondeeplearningnetworkforrefinedsegmentationofcracksfromshieldtunnelliningimages AT daoyuantan hybridattentiondeeplearningnetworkforrefinedsegmentationofcracksfromshieldtunnelliningimages AT hongweihuang hybridattentiondeeplearningnetworkforrefinedsegmentationofcracksfromshieldtunnelliningimages |