Integration of Vehicle Target Detection and Recognition in Large-Scene SAR Images Based on YOLOv5

In response to the problems existing in traditional SAR image target detection methods, such as complicated processes, long detection times, and poor detection effects in complex backgrounds, this paper proposes an integration of detection and recognition method based on deep learning for large-scene SAR images with vehicle targets. The paper introduces the issues of target identification through three stages of target detection, identification, and classification in traditional methods. To address these problems, this paper introduces a one-stage detection network based on YOLOv5 to construct a SAR image vehicle target detection and recognition algorithm. To verify the performance of the algorithm, this paper generated a dataset containing 10 different vehicle targets in large-scene SAR images and applied it to experiments. The results demonstrate that the algorithm has good performance and fast detection speed. The research results of this paper can provide important references for large-scale SAR image target detection.