Two-Stage Underwater Object Detection Network Using Swin Transformer
Underwater object detection plays an essential role in ocean exploration, and the increasing amount of underwater object image data makes the study of advanced underwater object detection algorithms of great practical significance. However, there are problems with colour offset, low contrast, and ta...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9938441/ |
| _version_ | 1828130308427350016 |
|---|---|
| author | Jia Liu Shuang Liu Shujuan Xu Changjun Zhou |
| author_facet | Jia Liu Shuang Liu Shujuan Xu Changjun Zhou |
| author_sort | Jia Liu |
| collection | DOAJ |
| description | Underwater object detection plays an essential role in ocean exploration, and the increasing amount of underwater object image data makes the study of advanced underwater object detection algorithms of great practical significance. However, there are problems with colour offset, low contrast, and target blur in underwater image data. An underwater object detection algorithm based on Faster R-CNN is proposed to solve these problems. First, the Swin Transformer is used as the backbone network of the algorithm. Second, by adding the path aggregation network, the deep feature map and the shallow feature map are superimposed and fused. Third, online hard example mining, makes the training process more efficient. Fourth, the ROI pooling is improved to ROI align, eliminating the two quantization errors of ROI pooling and improving the detection performance. Compared with other algorithms, the proposed algorithm’s based on improved Faster-RCNN on URPC2018 dataset is improved to 80.54%, and basically solve the problem of missed detection and false detection of objects of different sizes in a complex environment. |
| first_indexed | 2024-04-11T16:32:10Z |
| format | Article |
| id | doaj.art-060ed1ec8b0e425eb8e4aa4d4c3ae239 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-11T16:32:10Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-060ed1ec8b0e425eb8e4aa4d4c3ae2392022-12-22T04:13:59ZengIEEEIEEE Access2169-35362022-01-011011723511724710.1109/ACCESS.2022.32195929938441Two-Stage Underwater Object Detection Network Using Swin TransformerJia Liu0https://orcid.org/0000-0001-8598-8019Shuang Liu1https://orcid.org/0000-0002-0095-4328Shujuan Xu2https://orcid.org/0000-0001-5815-457XChangjun Zhou3https://orcid.org/0000-0002-0129-2231School of Computer Science and Engineering, Dalian Minzu University, Dalian, Liaoning, ChinaSchool of Computer Science and Engineering, Dalian Minzu University, Dalian, Liaoning, ChinaSchool of Computer Science and Engineering, Dalian Minzu University, Dalian, Liaoning, ChinaCollege of Mathematics and Computer Science, Zhejiang Normal University, Zhejiang, Jinhua, ChinaUnderwater object detection plays an essential role in ocean exploration, and the increasing amount of underwater object image data makes the study of advanced underwater object detection algorithms of great practical significance. However, there are problems with colour offset, low contrast, and target blur in underwater image data. An underwater object detection algorithm based on Faster R-CNN is proposed to solve these problems. First, the Swin Transformer is used as the backbone network of the algorithm. Second, by adding the path aggregation network, the deep feature map and the shallow feature map are superimposed and fused. Third, online hard example mining, makes the training process more efficient. Fourth, the ROI pooling is improved to ROI align, eliminating the two quantization errors of ROI pooling and improving the detection performance. Compared with other algorithms, the proposed algorithm’s based on improved Faster-RCNN on URPC2018 dataset is improved to 80.54%, and basically solve the problem of missed detection and false detection of objects of different sizes in a complex environment.https://ieeexplore.ieee.org/document/9938441/Deep learningunderwater target detectionswin transformerfeature fusionfaster R-CNNOHEM |
| spellingShingle | Jia Liu Shuang Liu Shujuan Xu Changjun Zhou Two-Stage Underwater Object Detection Network Using Swin Transformer IEEE Access Deep learning underwater target detection swin transformer feature fusion faster R-CNN OHEM |
| title | Two-Stage Underwater Object Detection Network Using Swin Transformer |
| title_full | Two-Stage Underwater Object Detection Network Using Swin Transformer |
| title_fullStr | Two-Stage Underwater Object Detection Network Using Swin Transformer |
| title_full_unstemmed | Two-Stage Underwater Object Detection Network Using Swin Transformer |
| title_short | Two-Stage Underwater Object Detection Network Using Swin Transformer |
| title_sort | two stage underwater object detection network using swin transformer |
| topic | Deep learning underwater target detection swin transformer feature fusion faster R-CNN OHEM |
| url | https://ieeexplore.ieee.org/document/9938441/ |
| work_keys_str_mv | AT jialiu twostageunderwaterobjectdetectionnetworkusingswintransformer AT shuangliu twostageunderwaterobjectdetectionnetworkusingswintransformer AT shujuanxu twostageunderwaterobjectdetectionnetworkusingswintransformer AT changjunzhou twostageunderwaterobjectdetectionnetworkusingswintransformer |