In the maritime ship detection of synthetic aperture radar (SAR) images, for the small ship targets in SAR images, the number of ships is large and the near-shore ship targets are affected by the complex background of the coast, which leads to wrong detection, omission and low recall in the ship detection results, this paper proposes an algorithmic model of ship target detection algorithm for SAR images, YOLOv5-MSX.The results of testing the improved YOLOv5-MSX model on the publicly available The test results on the SAR image ship dataset show that the mAP_0.5:0.95 value and recall of the improved YOLOv5-MSX model reach 71.7% and 89.8%, respectively, which are improved by 2.5% and 2% compared to the original YOLOv5s model, and the algorithm has a good robustness, which improves the overall effect of SAR ship image detection and recognition.
Research on SAR maritime vessel detection algorithm based on improved YOLOv5A YOLOv5-MSX algorithm for SAR ship image detection
2025-04-18
1423331 byte
Conference paper
Electronic Resource
English
STD-Yolov5: a ship-type detection model based on improved Yolov5
Taylor & Francis Verlag | 2024
|Mixed Traffic Flow Detection Algorithm Based on Improved YOLOv5
Springer Verlag | 2025
|