Ship Recognition Algorithm Based on ResNet in SAR Images

In the application fields of ocean target recognition in remote sensing images, the target classification of the marine ships based on synthetic aperture radar (SAR) figures remains a significant challenge. The traditional ship target recognition algorithms rely on manually selected features, and these features need to be designed on many experimental bases and professional domain knowledge, which leads to the poor robustness of the algorithm and poor recognition results. In this paper, in order to solve the problem of ship recognition in the SAR image without using manually selected features, a method based on the ResNet is proposed. First, a data augmentation module has been used to expand the experimental dataset. Then, the ResNet is used to recognize the ship in the SAR figures. Ultimately, the experiments based on the ship SAR dataset are carried out, and the suggested recognition method is verified to be of great effectiveness and applicability.