LRA-UNet: A Lightweight Residual Attention Network for SAR Marine Oil Spill Detection
Oil spills represent a serious threat to marine ecosystems. Remote sensing monitoring, especially based on synthetic aperture radar (SAR), have been extensively employed in marine environments due to its unique advantages. However, SAR images of marine oil spills exhibit challenges of weak boundarie...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-06-01
|
| Series: | Journal of Marine Science and Engineering |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2077-1312/13/6/1161 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839653648787832832 |
|---|---|
| author | Yu Cai Jingjing Su Jun Song Dekai Xu Liankang Zhang Gaoyuan Shen |
| author_facet | Yu Cai Jingjing Su Jun Song Dekai Xu Liankang Zhang Gaoyuan Shen |
| author_sort | Yu Cai |
| collection | DOAJ |
| description | Oil spills represent a serious threat to marine ecosystems. Remote sensing monitoring, especially based on synthetic aperture radar (SAR), have been extensively employed in marine environments due to its unique advantages. However, SAR images of marine oil spills exhibit challenges of weak boundaries, confusion with look-alike phenomena, and the difficulty of detecting small-scale targets. To address these issues, we propose LRA-UNet, a Lightweight Residual Attention UNet for semantic segmentation in SAR images. Our model integrates depthwise separable convolutions to reduce feature redundancy and computational cost, while adopting a residual encoder enhanced with the Simple Attention Module (SimAM) to improve the precise extraction of target features. Additionally, we design a joint loss function that incorporates Sobel-based edge information, emphasizing boundary features during training to enhance edge sharpness. Experimental results show that LRA-UNet achieves superior segmentation results, with a mIoU of 67.36%, surpassing the original UNet by 4.41%, and a 5.17% improvement in IoU for the oil spill category. These results confirm the effectiveness of our approach in accurately extracting oil spill regions from complex SAR imagery. |
| format | Article |
| id | doaj-art-fb80eb9e70d04e9896bae8ab036d2be7 |
| institution | Matheson Library |
| issn | 2077-1312 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Marine Science and Engineering |
| spelling | doaj-art-fb80eb9e70d04e9896bae8ab036d2be72025-06-25T14:01:38ZengMDPI AGJournal of Marine Science and Engineering2077-13122025-06-01136116110.3390/jmse13061161LRA-UNet: A Lightweight Residual Attention Network for SAR Marine Oil Spill DetectionYu Cai0Jingjing Su1Jun Song2Dekai Xu3Liankang Zhang4Gaoyuan Shen5Operational Oceanographic Institution, Dalian Ocean University, Dalian 116023, ChinaOperational Oceanographic Institution, Dalian Ocean University, Dalian 116023, ChinaOperational Oceanographic Institution, Dalian Ocean University, Dalian 116023, ChinaOperational Oceanographic Institution, Dalian Ocean University, Dalian 116023, ChinaOperational Oceanographic Institution, Dalian Ocean University, Dalian 116023, ChinaOperational Oceanographic Institution, Dalian Ocean University, Dalian 116023, ChinaOil spills represent a serious threat to marine ecosystems. Remote sensing monitoring, especially based on synthetic aperture radar (SAR), have been extensively employed in marine environments due to its unique advantages. However, SAR images of marine oil spills exhibit challenges of weak boundaries, confusion with look-alike phenomena, and the difficulty of detecting small-scale targets. To address these issues, we propose LRA-UNet, a Lightweight Residual Attention UNet for semantic segmentation in SAR images. Our model integrates depthwise separable convolutions to reduce feature redundancy and computational cost, while adopting a residual encoder enhanced with the Simple Attention Module (SimAM) to improve the precise extraction of target features. Additionally, we design a joint loss function that incorporates Sobel-based edge information, emphasizing boundary features during training to enhance edge sharpness. Experimental results show that LRA-UNet achieves superior segmentation results, with a mIoU of 67.36%, surpassing the original UNet by 4.41%, and a 5.17% improvement in IoU for the oil spill category. These results confirm the effectiveness of our approach in accurately extracting oil spill regions from complex SAR imagery.https://www.mdpi.com/2077-1312/13/6/1161oil spill detectionSAR imagesdeep learning modelimage segmentation |
| spellingShingle | Yu Cai Jingjing Su Jun Song Dekai Xu Liankang Zhang Gaoyuan Shen LRA-UNet: A Lightweight Residual Attention Network for SAR Marine Oil Spill Detection Journal of Marine Science and Engineering oil spill detection SAR images deep learning model image segmentation |
| title | LRA-UNet: A Lightweight Residual Attention Network for SAR Marine Oil Spill Detection |
| title_full | LRA-UNet: A Lightweight Residual Attention Network for SAR Marine Oil Spill Detection |
| title_fullStr | LRA-UNet: A Lightweight Residual Attention Network for SAR Marine Oil Spill Detection |
| title_full_unstemmed | LRA-UNet: A Lightweight Residual Attention Network for SAR Marine Oil Spill Detection |
| title_short | LRA-UNet: A Lightweight Residual Attention Network for SAR Marine Oil Spill Detection |
| title_sort | lra unet a lightweight residual attention network for sar marine oil spill detection |
| topic | oil spill detection SAR images deep learning model image segmentation |
| url | https://www.mdpi.com/2077-1312/13/6/1161 |
| work_keys_str_mv | AT yucai lraunetalightweightresidualattentionnetworkforsarmarineoilspilldetection AT jingjingsu lraunetalightweightresidualattentionnetworkforsarmarineoilspilldetection AT junsong lraunetalightweightresidualattentionnetworkforsarmarineoilspilldetection AT dekaixu lraunetalightweightresidualattentionnetworkforsarmarineoilspilldetection AT liankangzhang lraunetalightweightresidualattentionnetworkforsarmarineoilspilldetection AT gaoyuanshen lraunetalightweightresidualattentionnetworkforsarmarineoilspilldetection |