A Novel Detection Method Using YOLOv5 for Vehicle Target under Complex Situation

Aiming at the problem of low accuracy vehicle image detection and recognition caused by visibility in foggy weather, an improved YOLOv5 algorithm is proposed. This adjusts brightness contrast adding adaptive histogram equalization method to preprocessing, highlights detailed information signs, changes backbone network standard convolution mode depth separable for model lightweight processing. By constructing corresponding target data set, this paper superior object commonly used on public set terms performance effectiveness, draws following conclusion from comparison results ablation experiments, improves a single reduces processing time.