Road Crack Detection Algorithm Based on Improved YOLOV5 Model

A lightweight road crack target detection method based on YOLOv5 and GhostNet is proposed to address the problems of complex models and high computational complexity in existing road crack detection methods. The GhostConv and GhostC3 modules of the lightweight network GhostNet are introduced to improve the backbone network of YOLOv5, achieving the goal of significantly reducing model computation while ensuring the network's feature extraction capability. The experimental results show that the improved algorithm proposed in this paper significantly reduces the volume of the model and the number of parameters, and at the same time ensures the high precision of detection. Therefore, the model can be used as an ideal lightweight detection model for road crack images.