ESD-YOLO11: High-Precision Multi-Scale and Small Target Ship Detection in SAR Imagery

Synthetic aperture radar (SAR) image ship detection plays a vital role in fields such as marine monitoring. However, ship detection still faces great challenges due to large changes in target scale, and difficulties in detecting small targets. To address these issues, this paper proposes a SAR ship detection method based on improved YOLO11, namely ESD-YOLO11. First, the EMA attention mechanism is introduced at the end of the backbone network to capture global contextual information through cross-space learning, and combined with multi-scale feature fusion to improve the model's adaptability to complex scenes. Then, the SCConv module is integrated into the C3k2 network structure to obtain better performance by separating redundant features and dynamically reconstructing channel and spatial information. Finally, a multi-dimensional dynamic adaptive detection head Dynamic Head is introduced. Experimental results on the HRSID and SSDD datasets show that the method achieves 94.7% and 98.6% in the mAP50 metric, respectively, an increase of 1.9% and 1.2% over the original YOLO11 model, verifying the good performance of ESD-YOLO11 in the SAR ship detection task.