Dam Crack Image Detection Model on Feature Enhancement and Attention Mechanism
Dam crack detection can effectively avoid safety accidents of dams. To solve the problem that the dam crack image samples are not available and the traditional algorithm detects cracks with low accuracy, we provide a dam crack image detection model based on crack feature enhancement and attention me...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-12-01
|
| Series: | Water |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4441/15/1/64 |
| _version_ | 1797431089370759168 |
|---|---|
| author | Guoyan Xu Xu Han Yuwei Zhang Chunyan Wu |
| author_facet | Guoyan Xu Xu Han Yuwei Zhang Chunyan Wu |
| author_sort | Guoyan Xu |
| collection | DOAJ |
| description | Dam crack detection can effectively avoid safety accidents of dams. To solve the problem that the dam crack image samples are not available and the traditional algorithm detects cracks with low accuracy, we provide a dam crack image detection model based on crack feature enhancement and attention mechanism. Firstly, we expand the dam crack image dataset through a generative adversarial network based on crack feature enhancement (Cracks Enhancements GAN, CE-GAN). It can fully expand the dam crack data samples and improve the quality of the training data. Secondly, we propose a crack image detection model based on the attention mechanism (Attention-based Faster-RCNN, AF-RCNN). The attention mechanism is added in the crack detection module to give different weights to the proposal boxes around the crack target and fuse the candidate boxes with high weights to accurately detect the crack target location. The experimental results show that our algorithm achieves 81.07% mAP on the expanded dam crack dataset, which is 8.39% higher than the original Faster-RCNN algorithm. The detection accuracy is significantly improved compared with other traditional dam crack detection algorithm models. |
| first_indexed | 2024-03-09T09:37:55Z |
| format | Article |
| id | doaj.art-8d8bb1e6623d4b8b90637444f95b103d |
| institution | Directory Open Access Journal |
| issn | 2073-4441 |
| language | English |
| last_indexed | 2024-03-09T09:37:55Z |
| publishDate | 2022-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Water |
| spelling | doaj.art-8d8bb1e6623d4b8b90637444f95b103d2023-12-02T01:13:45ZengMDPI AGWater2073-44412022-12-011516410.3390/w15010064Dam Crack Image Detection Model on Feature Enhancement and Attention MechanismGuoyan Xu0Xu Han1Yuwei Zhang2Chunyan Wu3College of Computer and Information, Hohai University, Nanjing 211100, ChinaCollege of Computer and Information, Hohai University, Nanjing 211100, ChinaCollege of Computer and Information, Hohai University, Nanjing 211100, ChinaCollege of Computer and Information, Hohai University, Nanjing 211100, ChinaDam crack detection can effectively avoid safety accidents of dams. To solve the problem that the dam crack image samples are not available and the traditional algorithm detects cracks with low accuracy, we provide a dam crack image detection model based on crack feature enhancement and attention mechanism. Firstly, we expand the dam crack image dataset through a generative adversarial network based on crack feature enhancement (Cracks Enhancements GAN, CE-GAN). It can fully expand the dam crack data samples and improve the quality of the training data. Secondly, we propose a crack image detection model based on the attention mechanism (Attention-based Faster-RCNN, AF-RCNN). The attention mechanism is added in the crack detection module to give different weights to the proposal boxes around the crack target and fuse the candidate boxes with high weights to accurately detect the crack target location. The experimental results show that our algorithm achieves 81.07% mAP on the expanded dam crack dataset, which is 8.39% higher than the original Faster-RCNN algorithm. The detection accuracy is significantly improved compared with other traditional dam crack detection algorithm models.https://www.mdpi.com/2073-4441/15/1/64dam crack detectionFSLGANattention mechanismfeature enhancement |
| spellingShingle | Guoyan Xu Xu Han Yuwei Zhang Chunyan Wu Dam Crack Image Detection Model on Feature Enhancement and Attention Mechanism Water dam crack detection FSL GAN attention mechanism feature enhancement |
| title | Dam Crack Image Detection Model on Feature Enhancement and Attention Mechanism |
| title_full | Dam Crack Image Detection Model on Feature Enhancement and Attention Mechanism |
| title_fullStr | Dam Crack Image Detection Model on Feature Enhancement and Attention Mechanism |
| title_full_unstemmed | Dam Crack Image Detection Model on Feature Enhancement and Attention Mechanism |
| title_short | Dam Crack Image Detection Model on Feature Enhancement and Attention Mechanism |
| title_sort | dam crack image detection model on feature enhancement and attention mechanism |
| topic | dam crack detection FSL GAN attention mechanism feature enhancement |
| url | https://www.mdpi.com/2073-4441/15/1/64 |
| work_keys_str_mv | AT guoyanxu damcrackimagedetectionmodelonfeatureenhancementandattentionmechanism AT xuhan damcrackimagedetectionmodelonfeatureenhancementandattentionmechanism AT yuweizhang damcrackimagedetectionmodelonfeatureenhancementandattentionmechanism AT chunyanwu damcrackimagedetectionmodelonfeatureenhancementandattentionmechanism |