AJANet: SAR Ship Detection Network Based on Adaptive Channel Attention and Large Separable Kernel Adaptation
Due to issues such as low resolution, scattering noise, and background clutter, ship detection in Synthetic Aperture Radar (SAR) images remains challenging, especially in inshore regions, where these factors have similar scattering characteristics. To overcome these challenges, this paper proposes a...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-05-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/17/10/1745 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849327102546935808 |
|---|---|
| author | Yishuang Chen Jie Chen Long Sun Bocai Wu Hui Xu |
| author_facet | Yishuang Chen Jie Chen Long Sun Bocai Wu Hui Xu |
| author_sort | Yishuang Chen |
| collection | DOAJ |
| description | Due to issues such as low resolution, scattering noise, and background clutter, ship detection in Synthetic Aperture Radar (SAR) images remains challenging, especially in inshore regions, where these factors have similar scattering characteristics. To overcome these challenges, this paper proposes a novel SAR ship detection framework that integrates adaptive channel attention with large kernel adaptation. The proposed method improves multi-scale contextual information extraction by enhancing feature map interactions at different scales. This method effectively reduces false positives, missed detections, and localization ambiguities, especially in complex inshore environments. Also, it includes an adaptive channel attention block that adjusts attention weights according to the dimensions of the input feature maps, enabling the model to prioritize local information and improve sensitivity to small object features in SAR images. In addition, a large kernel attention block with adaptive kernel size is introduced to automatically adjust the receptive field designed to extract abundant context information at different detection layers. Experimental evaluations on the SSDD and Hysid SAR ship datasets indicate that our method achieves excellent detection performance compared to current methods, as well as demonstrate its effectiveness in overcoming SAR ship detection challenges. |
| format | Article |
| id | doaj-art-87f4d9b2e4bf471ebcf7224874341aca |
| institution | Kabale University |
| issn | 2072-4292 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj-art-87f4d9b2e4bf471ebcf7224874341aca2025-08-20T03:47:58ZengMDPI AGRemote Sensing2072-42922025-05-011710174510.3390/rs17101745AJANet: SAR Ship Detection Network Based on Adaptive Channel Attention and Large Separable Kernel AdaptationYishuang Chen0Jie Chen1Long Sun2Bocai Wu3Hui Xu4East China Research Institute of Electronic Engineering, Hefei 230088, ChinaKey Laboratory of Intelligent Computing and Signal Processing of Ministry of Education, School of Electronics and Information Engineering, Anhui University, Hefei 230601, ChinaANHUI SUN CREATE ELECTRONICS Co., Ltd., Hefei 230031, ChinaEast China Research Institute of Electronic Engineering, Hefei 230088, ChinaKey Laboratory of Intelligent Computing and Signal Processing of Ministry of Education, School of Electronics and Information Engineering, Anhui University, Hefei 230601, ChinaDue to issues such as low resolution, scattering noise, and background clutter, ship detection in Synthetic Aperture Radar (SAR) images remains challenging, especially in inshore regions, where these factors have similar scattering characteristics. To overcome these challenges, this paper proposes a novel SAR ship detection framework that integrates adaptive channel attention with large kernel adaptation. The proposed method improves multi-scale contextual information extraction by enhancing feature map interactions at different scales. This method effectively reduces false positives, missed detections, and localization ambiguities, especially in complex inshore environments. Also, it includes an adaptive channel attention block that adjusts attention weights according to the dimensions of the input feature maps, enabling the model to prioritize local information and improve sensitivity to small object features in SAR images. In addition, a large kernel attention block with adaptive kernel size is introduced to automatically adjust the receptive field designed to extract abundant context information at different detection layers. Experimental evaluations on the SSDD and Hysid SAR ship datasets indicate that our method achieves excellent detection performance compared to current methods, as well as demonstrate its effectiveness in overcoming SAR ship detection challenges.https://www.mdpi.com/2072-4292/17/10/1745SAR ship detectionadaptive channel attentionadaptive large kernel attention |
| spellingShingle | Yishuang Chen Jie Chen Long Sun Bocai Wu Hui Xu AJANet: SAR Ship Detection Network Based on Adaptive Channel Attention and Large Separable Kernel Adaptation Remote Sensing SAR ship detection adaptive channel attention adaptive large kernel attention |
| title | AJANet: SAR Ship Detection Network Based on Adaptive Channel Attention and Large Separable Kernel Adaptation |
| title_full | AJANet: SAR Ship Detection Network Based on Adaptive Channel Attention and Large Separable Kernel Adaptation |
| title_fullStr | AJANet: SAR Ship Detection Network Based on Adaptive Channel Attention and Large Separable Kernel Adaptation |
| title_full_unstemmed | AJANet: SAR Ship Detection Network Based on Adaptive Channel Attention and Large Separable Kernel Adaptation |
| title_short | AJANet: SAR Ship Detection Network Based on Adaptive Channel Attention and Large Separable Kernel Adaptation |
| title_sort | ajanet sar ship detection network based on adaptive channel attention and large separable kernel adaptation |
| topic | SAR ship detection adaptive channel attention adaptive large kernel attention |
| url | https://www.mdpi.com/2072-4292/17/10/1745 |
| work_keys_str_mv | AT yishuangchen ajanetsarshipdetectionnetworkbasedonadaptivechannelattentionandlargeseparablekerneladaptation AT jiechen ajanetsarshipdetectionnetworkbasedonadaptivechannelattentionandlargeseparablekerneladaptation AT longsun ajanetsarshipdetectionnetworkbasedonadaptivechannelattentionandlargeseparablekerneladaptation AT bocaiwu ajanetsarshipdetectionnetworkbasedonadaptivechannelattentionandlargeseparablekerneladaptation AT huixu ajanetsarshipdetectionnetworkbasedonadaptivechannelattentionandlargeseparablekerneladaptation |