In the detection of synthetic aperture radar (SAR) image ship targets, the traditional artificial feature-based target detection method is less effective due to the complex background. Based on the single-stage target detection algorithm RetinaNet in deep learning, this paper combines the characteristics of SAR image with less feature information, adopts the idea of multi-feature layer fusion and proposes a more appropriate loss function calculation method. Then we use the SSDD data set to train the network, and improve the robustness and convergence speed of the algorithm through data augmentation and transfer Learning. Finally, we compare other target detection algorithms based on deep learning through experiments. The results show that our method has higher detection accuracy.