An Anchor-Free Method Based on Transformers and Adaptive Features for Arbitrarily Oriented Ship Detection in SAR Images
Ship detection is a crucial application of synthetic aperture radar (SAR). Most recent studies have relied on convolutional neural networks (CNNs). CNNs tend to struggle in gathering adequate contextual information through local receptive fields and are also susceptible to noise. Inshore scenes in S...
| 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/10287400/ |
| _version_ | 1797372221880008704 |
|---|---|
| author | Bingji Chen Chunrui Yu Shuang Zhao Hongjun Song |
| author_facet | Bingji Chen Chunrui Yu Shuang Zhao Hongjun Song |
| author_sort | Bingji Chen |
| collection | DOAJ |
| description | Ship detection is a crucial application of synthetic aperture radar (SAR). Most recent studies have relied on convolutional neural networks (CNNs). CNNs tend to struggle in gathering adequate contextual information through local receptive fields and are also susceptible to noise. Inshore scenes in SAR images are plagued by substantial background noise, so achieving high-accuracy ship detection of arbitrary orientations within complex scenes remains an ongoing challenge when relying solely on CNNs. To address the above challenges, this article presents an anchor-free method based on transformers and adaptive features, namely, SAD-Det, which can detect rotationally invariant ship targets with high average precision in SAR images. Specifically, a transformer-based backbone network called the ship spatial pooling pyramid vision transformer is proposed to enhance the long-range dependencies and obtain sufficient contextual information for ships in SAR images. In addition, a neck network called the adaptive feature pyramid network is designed to enhance the ability of ship feature adaptation by adding fusion factors to feature layers in SAR images. Finally, a head network called the deformable head is constructed to make the network more adaptable to the characteristics of ships by adaptively detecting the spatial sampling positions of the targets in SAR images. The effectiveness of the proposed method is verified by experiments on two publicly available datasets, i.e., SAR ship detection dataset and rotated ship detection dataset in SAR images. Compared with other arbitrarily oriented object detection methods, the proposed method achieves state-of-the-art detection performance. |
| first_indexed | 2024-03-08T18:31:41Z |
| format | Article |
| id | doaj.art-5c243c062e25449e9caaef3353d868a0 |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-03-08T18:31:41Z |
| 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-5c243c062e25449e9caaef3353d868a02023-12-30T00:00:59ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352024-01-01172012202810.1109/JSTARS.2023.332557310287400An Anchor-Free Method Based on Transformers and Adaptive Features for Arbitrarily Oriented Ship Detection in SAR ImagesBingji Chen0https://orcid.org/0000-0003-0023-4854Chunrui Yu1Shuang Zhao2https://orcid.org/0000-0002-9370-4183Hongjun Song3https://orcid.org/0009-0008-0096-3364Department of Space Microwave Remote Sensing System, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaBeijing Institute of Tracking and Telecommunication Technology, Beijing, ChinaDepartment of Space Microwave Remote Sensing System, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaDepartment of Space Microwave Remote Sensing System, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaShip detection is a crucial application of synthetic aperture radar (SAR). Most recent studies have relied on convolutional neural networks (CNNs). CNNs tend to struggle in gathering adequate contextual information through local receptive fields and are also susceptible to noise. Inshore scenes in SAR images are plagued by substantial background noise, so achieving high-accuracy ship detection of arbitrary orientations within complex scenes remains an ongoing challenge when relying solely on CNNs. To address the above challenges, this article presents an anchor-free method based on transformers and adaptive features, namely, SAD-Det, which can detect rotationally invariant ship targets with high average precision in SAR images. Specifically, a transformer-based backbone network called the ship spatial pooling pyramid vision transformer is proposed to enhance the long-range dependencies and obtain sufficient contextual information for ships in SAR images. In addition, a neck network called the adaptive feature pyramid network is designed to enhance the ability of ship feature adaptation by adding fusion factors to feature layers in SAR images. Finally, a head network called the deformable head is constructed to make the network more adaptable to the characteristics of ships by adaptively detecting the spatial sampling positions of the targets in SAR images. The effectiveness of the proposed method is verified by experiments on two publicly available datasets, i.e., SAR ship detection dataset and rotated ship detection dataset in SAR images. Compared with other arbitrarily oriented object detection methods, the proposed method achieves state-of-the-art detection performance.https://ieeexplore.ieee.org/document/10287400/Adaptive featuresanchor-freearbitrarily oriented detectorship detectionsynthetic aperture radar (SAR)transformer |
| spellingShingle | Bingji Chen Chunrui Yu Shuang Zhao Hongjun Song An Anchor-Free Method Based on Transformers and Adaptive Features for Arbitrarily Oriented Ship Detection in SAR Images IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Adaptive features anchor-free arbitrarily oriented detector ship detection synthetic aperture radar (SAR) transformer |
| title | An Anchor-Free Method Based on Transformers and Adaptive Features for Arbitrarily Oriented Ship Detection in SAR Images |
| title_full | An Anchor-Free Method Based on Transformers and Adaptive Features for Arbitrarily Oriented Ship Detection in SAR Images |
| title_fullStr | An Anchor-Free Method Based on Transformers and Adaptive Features for Arbitrarily Oriented Ship Detection in SAR Images |
| title_full_unstemmed | An Anchor-Free Method Based on Transformers and Adaptive Features for Arbitrarily Oriented Ship Detection in SAR Images |
| title_short | An Anchor-Free Method Based on Transformers and Adaptive Features for Arbitrarily Oriented Ship Detection in SAR Images |
| title_sort | anchor free method based on transformers and adaptive features for arbitrarily oriented ship detection in sar images |
| topic | Adaptive features anchor-free arbitrarily oriented detector ship detection synthetic aperture radar (SAR) transformer |
| url | https://ieeexplore.ieee.org/document/10287400/ |
| work_keys_str_mv | AT bingjichen ananchorfreemethodbasedontransformersandadaptivefeaturesforarbitrarilyorientedshipdetectioninsarimages AT chunruiyu ananchorfreemethodbasedontransformersandadaptivefeaturesforarbitrarilyorientedshipdetectioninsarimages AT shuangzhao ananchorfreemethodbasedontransformersandadaptivefeaturesforarbitrarilyorientedshipdetectioninsarimages AT hongjunsong ananchorfreemethodbasedontransformersandadaptivefeaturesforarbitrarilyorientedshipdetectioninsarimages AT bingjichen anchorfreemethodbasedontransformersandadaptivefeaturesforarbitrarilyorientedshipdetectioninsarimages AT chunruiyu anchorfreemethodbasedontransformersandadaptivefeaturesforarbitrarilyorientedshipdetectioninsarimages AT shuangzhao anchorfreemethodbasedontransformersandadaptivefeaturesforarbitrarilyorientedshipdetectioninsarimages AT hongjunsong anchorfreemethodbasedontransformersandadaptivefeaturesforarbitrarilyorientedshipdetectioninsarimages |