Large Convolution Kernel Network With Edge Self-Attention for Oriented SAR Ship Detection
Synthetic aperture radar (SAR) ship detection is a popular area in remote sensing, which has broad applications in fishery management, maritime rescue and marine detection. In recent years, deep learning has been successfully applied in this field, but SAR ship detection still faces some challenges,...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10798974/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841557081302237184 |
|---|---|
| author | Yan Feng Yupeng Zhang Xiangqing Zhang Yuning Wang Shaohui Mei |
| author_facet | Yan Feng Yupeng Zhang Xiangqing Zhang Yuning Wang Shaohui Mei |
| author_sort | Yan Feng |
| collection | DOAJ |
| description | Synthetic aperture radar (SAR) ship detection is a popular area in remote sensing, which has broad applications in fishery management, maritime rescue and marine detection. In recent years, deep learning has been successfully applied in this field, but SAR ship detection still faces some challenges, such as speckle noise, strong scattering interference from complex backgrounds, and insufficient use of SAR image characteristics. These issues result in inaccurate ship positioning in complex scenes and a high rate of false alarms. To address the above challenges, we propose a large convolution kernel network with edge self-attention for oriented SAR ship detection (LKE-Det), which can accurately detect arbitrarily rotated ship targets in SAR images. First, we propose a backbone network based on large convolutional kernels, which has a large receptive field and can enhance the long-range dependencies of ships. The backbone adopts a bifurcated gate unit structure to combine the multiscale parallel convolution kernel module and the large convolution kernel attention module. The structure captures feature information of different scales while obtaining sufficient context information. Second, we adopt an edge self-attention mechanism, which uses an edge detection algorithm to extract the gradient matrix of the image and integrate it into the feature pyramid structure to effectively enhance the contour information of ships. The extensive experiments on SSDD and RSDD-SAR datasets demonstrate that our LKE-Det outperforms the state-of-the-art methods. |
| format | Article |
| id | doaj-art-44c85f461e0d4d6ab910991a6ce619b8 |
| institution | Kabale University |
| issn | 1939-1404 2151-1535 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj-art-44c85f461e0d4d6ab910991a6ce619b82025-01-07T00:00:42ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-01182867287910.1109/JSTARS.2024.351485510798974Large Convolution Kernel Network With Edge Self-Attention for Oriented SAR Ship DetectionYan Feng0https://orcid.org/0000-0002-0669-9970Yupeng Zhang1https://orcid.org/0009-0009-3015-9649Xiangqing Zhang2https://orcid.org/0000-0001-7273-6170Yuning Wang3https://orcid.org/0009-0001-8222-9161Shaohui Mei4https://orcid.org/0000-0002-8018-596XSchool of Electronics and Information, Northwestern Polytechnical University, Xi'an, ChinaSchool of Electronics and Information, Northwestern Polytechnical University, Xi'an, ChinaSchool of Electronics and Information, Northwestern Polytechnical University, Xi'an, ChinaSchool of Electronics and Information, Northwestern Polytechnical University, Xi'an, ChinaSchool of Electronics and Information, Northwestern Polytechnical University, Xi'an, ChinaSynthetic aperture radar (SAR) ship detection is a popular area in remote sensing, which has broad applications in fishery management, maritime rescue and marine detection. In recent years, deep learning has been successfully applied in this field, but SAR ship detection still faces some challenges, such as speckle noise, strong scattering interference from complex backgrounds, and insufficient use of SAR image characteristics. These issues result in inaccurate ship positioning in complex scenes and a high rate of false alarms. To address the above challenges, we propose a large convolution kernel network with edge self-attention for oriented SAR ship detection (LKE-Det), which can accurately detect arbitrarily rotated ship targets in SAR images. First, we propose a backbone network based on large convolutional kernels, which has a large receptive field and can enhance the long-range dependencies of ships. The backbone adopts a bifurcated gate unit structure to combine the multiscale parallel convolution kernel module and the large convolution kernel attention module. The structure captures feature information of different scales while obtaining sufficient context information. Second, we adopt an edge self-attention mechanism, which uses an edge detection algorithm to extract the gradient matrix of the image and integrate it into the feature pyramid structure to effectively enhance the contour information of ships. The extensive experiments on SSDD and RSDD-SAR datasets demonstrate that our LKE-Det outperforms the state-of-the-art methods.https://ieeexplore.ieee.org/document/10798974/Deep learningedge self-attentionlarge convolution kerneloriented ship detectionsynthetic aperture radar (SAR) |
| spellingShingle | Yan Feng Yupeng Zhang Xiangqing Zhang Yuning Wang Shaohui Mei Large Convolution Kernel Network With Edge Self-Attention for Oriented SAR Ship Detection IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Deep learning edge self-attention large convolution kernel oriented ship detection synthetic aperture radar (SAR) |
| title | Large Convolution Kernel Network With Edge Self-Attention for Oriented SAR Ship Detection |
| title_full | Large Convolution Kernel Network With Edge Self-Attention for Oriented SAR Ship Detection |
| title_fullStr | Large Convolution Kernel Network With Edge Self-Attention for Oriented SAR Ship Detection |
| title_full_unstemmed | Large Convolution Kernel Network With Edge Self-Attention for Oriented SAR Ship Detection |
| title_short | Large Convolution Kernel Network With Edge Self-Attention for Oriented SAR Ship Detection |
| title_sort | large convolution kernel network with edge self attention for oriented sar ship detection |
| topic | Deep learning edge self-attention large convolution kernel oriented ship detection synthetic aperture radar (SAR) |
| url | https://ieeexplore.ieee.org/document/10798974/ |
| work_keys_str_mv | AT yanfeng largeconvolutionkernelnetworkwithedgeselfattentionfororientedsarshipdetection AT yupengzhang largeconvolutionkernelnetworkwithedgeselfattentionfororientedsarshipdetection AT xiangqingzhang largeconvolutionkernelnetworkwithedgeselfattentionfororientedsarshipdetection AT yuningwang largeconvolutionkernelnetworkwithedgeselfattentionfororientedsarshipdetection AT shaohuimei largeconvolutionkernelnetworkwithedgeselfattentionfororientedsarshipdetection |