SAR Image Ship Detection Algorithm based on Machine Vision YOLO Model

This paper mainly studies the detection technology of tiny size ship targets in SAR images based on the deep learning method. Firstly, the feature extraction network structure is redesigned to solve the problem that the feature information of small and medium-sized targets in SAR images is easily lost. Then, the path enhancement method is used to enrich the geometric position information in the high-level feature graph, and the fusion of adjacent feature graphs is used to enrich the expression ability of the feature graph. Finally, the effectiveness of the proposed algorithm is verified by experiments, and the detection results in some typical scenarios are shown and compared with the classical target detection algorithm.