Lightweight SAR Ship Detection Network Based on Transformer and Feature Enhancement
With the rapid development of synthetic aperture radar (SAR) technology, SAR remote sensing has a wide range of applications in fields, such as marine surveillance and sea rescue. Currently, the SAR ship detection model based on deep learning suffers from the problems of low detection in real time a...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10423087/ |
| _version_ | 1797293283174514688 |
|---|---|
| author | Shichuang Zhou Ming Zhang Liang Wu Dahua Yu Jianjun Li Fei Fan Liyun Zhang Yang Liu |
| author_facet | Shichuang Zhou Ming Zhang Liang Wu Dahua Yu Jianjun Li Fei Fan Liyun Zhang Yang Liu |
| author_sort | Shichuang Zhou |
| collection | DOAJ |
| description | With the rapid development of synthetic aperture radar (SAR) technology, SAR remote sensing has a wide range of applications in fields, such as marine surveillance and sea rescue. Currently, the SAR ship detection model based on deep learning suffers from the problems of low detection in real time and low detection accuracy. In order to solve the abovementioned problems, this article proposes a lightweight SAR ship detection network (EGTB-Net) based on transformer and feature enhancement. First, we design a novel Ghost-ECA model as the backbone network of EGTB-Net, which reduces the number of parameters of the model and enhances the ability to identify key feature information at the same time. Then, we incorporate the transformer block in the backbone network to capture long-range dependencies, enrich contextual information, and improve the network's ability to capture different types of local information. Finally, we adopt a new SIoU loss function, which is used to solve the direction problem of mismatch between the real frame and the predicted frame and improve the network's ability to localize ship targets. The experimental results on the SAR-ship-dataset show that the mean average precision of the method is 94.83%, the detection speed is 61 frames per second, and the model size is only 5.94 M, while the model has excellent anti-interference ability. |
| first_indexed | 2024-03-07T20:10:40Z |
| format | Article |
| id | doaj.art-33662b69d6aa4c8fb210d6ac0dae066a |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-03-07T20:10:40Z |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-33662b69d6aa4c8fb210d6ac0dae066a2024-02-28T00:00:18ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352024-01-01174845485810.1109/JSTARS.2024.336295410423087Lightweight SAR Ship Detection Network Based on Transformer and Feature EnhancementShichuang Zhou0https://orcid.org/0009-0006-3936-468XMing Zhang1https://orcid.org/0000-0003-2638-9002Liang Wu2https://orcid.org/0000-0001-5988-4021Dahua Yu3https://orcid.org/0000-0001-7850-7512Jianjun Li4https://orcid.org/0000-0003-3003-8344Fei Fan5https://orcid.org/0009-0001-1748-5513Liyun Zhang6https://orcid.org/0009-0005-8292-2686Yang Liu7https://orcid.org/0009-0002-4737-8505School of Digital and Intelligence Industry, Inner Mongolia University of Science and Technology, Baotou, ChinaSchool of Digital and Intelligence Industry, Inner Mongolia University of Science and Technology, Baotou, ChinaSchool of Radiology, Shandong First Medical University and Shandong Academy of Medical Sciences, Tai'an, ChinaSchool of Digital and Intelligence Industry, Inner Mongolia University of Science and Technology, Baotou, ChinaSchool of Digital and Intelligence Industry, Inner Mongolia University of Science and Technology, Baotou, ChinaSchool of Digital and Intelligence Industry, Inner Mongolia University of Science and Technology, Baotou, ChinaSchool of Digital and Intelligence Industry, Inner Mongolia University of Science and Technology, Baotou, ChinaSchool of Digital and Intelligence Industry, Inner Mongolia University of Science and Technology, Baotou, ChinaWith the rapid development of synthetic aperture radar (SAR) technology, SAR remote sensing has a wide range of applications in fields, such as marine surveillance and sea rescue. Currently, the SAR ship detection model based on deep learning suffers from the problems of low detection in real time and low detection accuracy. In order to solve the abovementioned problems, this article proposes a lightweight SAR ship detection network (EGTB-Net) based on transformer and feature enhancement. First, we design a novel Ghost-ECA model as the backbone network of EGTB-Net, which reduces the number of parameters of the model and enhances the ability to identify key feature information at the same time. Then, we incorporate the transformer block in the backbone network to capture long-range dependencies, enrich contextual information, and improve the network's ability to capture different types of local information. Finally, we adopt a new SIoU loss function, which is used to solve the direction problem of mismatch between the real frame and the predicted frame and improve the network's ability to localize ship targets. The experimental results on the SAR-ship-dataset show that the mean average precision of the method is 94.83%, the detection speed is 61 frames per second, and the model size is only 5.94 M, while the model has excellent anti-interference ability.https://ieeexplore.ieee.org/document/10423087/Complex backgroundfeature enhancementlightweight networkship detectionsynthetic aperture radar (SAR) |
| spellingShingle | Shichuang Zhou Ming Zhang Liang Wu Dahua Yu Jianjun Li Fei Fan Liyun Zhang Yang Liu Lightweight SAR Ship Detection Network Based on Transformer and Feature Enhancement IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Complex background feature enhancement lightweight network ship detection synthetic aperture radar (SAR) |
| title | Lightweight SAR Ship Detection Network Based on Transformer and Feature Enhancement |
| title_full | Lightweight SAR Ship Detection Network Based on Transformer and Feature Enhancement |
| title_fullStr | Lightweight SAR Ship Detection Network Based on Transformer and Feature Enhancement |
| title_full_unstemmed | Lightweight SAR Ship Detection Network Based on Transformer and Feature Enhancement |
| title_short | Lightweight SAR Ship Detection Network Based on Transformer and Feature Enhancement |
| title_sort | lightweight sar ship detection network based on transformer and feature enhancement |
| topic | Complex background feature enhancement lightweight network ship detection synthetic aperture radar (SAR) |
| url | https://ieeexplore.ieee.org/document/10423087/ |
| work_keys_str_mv | AT shichuangzhou lightweightsarshipdetectionnetworkbasedontransformerandfeatureenhancement AT mingzhang lightweightsarshipdetectionnetworkbasedontransformerandfeatureenhancement AT liangwu lightweightsarshipdetectionnetworkbasedontransformerandfeatureenhancement AT dahuayu lightweightsarshipdetectionnetworkbasedontransformerandfeatureenhancement AT jianjunli lightweightsarshipdetectionnetworkbasedontransformerandfeatureenhancement AT feifan lightweightsarshipdetectionnetworkbasedontransformerandfeatureenhancement AT liyunzhang lightweightsarshipdetectionnetworkbasedontransformerandfeatureenhancement AT yangliu lightweightsarshipdetectionnetworkbasedontransformerandfeatureenhancement |