A Vehicle Recognition Model Based on Improved YOLOv5

The rapid development of the automobile industry has made life easier for people, but traffic accidents have increased in frequency recent years, making vehicle safety particularly important. This paper proposes an improved YOLOv5s algorithm identification and detection to reduce driving issues based on this problem. In order solve problems a disappearing model training gradient algorithm, difficulty recognizing small objects poor recognition accuracy caused by boundary frame regression function, it is necessary implement new function. These aspects been enhanced article. On basis traditional ELU activation function used replace original attention mechanism module then added algorithm’s backbone network improve feature extraction medium-sized objects. CIoU Loss replaces YOLOv5s, thereby enhancing convergence rate measurement precision loss paper, constructed dataset utilized conduct pertinent experiments. experimental results demonstrate that, compared previous mAP 3.1% higher, 0.8% 2.5% lower.