Multilevel Pyramid Feature Extraction and Task Decoupling Network for SAR Ship Detection

Synthetic Aperture Radar (SAR) target detection plays a crucial role in both military and civilian fields, attracting significant attention from researchers globally. CenterNet, a single-stage target detection method, is known for its high detection speed and accuracy by eliminating anchor-related calculations and non-maximum suppression (NMS). However, directly applying CenterNet to SAR ship detection poses challenges due to the distinctive characteristics of SAR images, including lower resolution, lower signal-to-noise ratio, and larger ship aspect ratios. To address these challenges, we propose MPDNet. which introduces a Multi-level Pyramid Feature Extraction Module (MP-FEM) to replace the encoding-decoding structure in CenterNet. MP-FEM employs multi-level pyramid and channel compression to fuse multi-scale SAR image features and acquire deep features quickly. Secondly, we propose the Convolution Channel Attention Module (Conv-CAM), which improves the multilayer perceptron in the common pooling attention mechanism into a multi-stage and one-dimension convolution. Therefore, the feature extraction capability of MP-FEM is further refined. Furthermore, we propose the Detection Task Decoupling Module (DTDM), which considers the characteristics of SAR ships and effectively detects smaller targets of different sizes, distinguishing the centers and sizes of densely arranged ships. DTDM extracts task-related features from the original feature map before inputting it into the three detection headers, thereby addressing the problem of task coupling in CenterNet's detection header module for SAR ship detection. Finally, the experimental results on SSDD dataset and SAR-ship-dataset show that the proposed network can significantly improve the SAR target detection accuracy.