YOLO-M: An Efficient YOLO Variant with MobileOne Backbone for Real-Time License Plate Detection

This study introduces the optimized YOLO-M model for real-time license plate detection. Leveraging performance metrics such as the F1-Confidence curve, Precision-Confidence curve, Precision-Recall curve, and Recall-Confidence curve, the model demonstrates exceptional precision and recall rates in real-time conditions. Specifically, YOLO-M achieves an optimal F1 score of 0.90 at a confidence threshold of 0.572, and reaches a peak precision of 1.00 at a confidence threshold of 0.983. Moreover, the model maintains a high precision of 0.925 at the mAP@0.5 recall level, and exhibits a high recall of 0.96 even at the lowest confidence threshold of 0.000. These findings underscore the reliability and practicality of YOLO-M for real-time license plate detection, providing significant technological support for traffic monitoring and security surveillance systems.