A steel surface defect detection algorithm based on improved YOLOv5

Abstract

In order to solve the problem of steel surface defect detection, an improved algorithm based on YOLOv5 is proposed. EIOU loss is used to replace the original GIOU loss function, and the attention mechanism SE module is added to the network model to strengthen important characteristic channels. By setting different training parameters in the steel defect set for multiple rounds of testing, the results show that under different parameters, the improved YOLOv5s model can detect steel surface defects with the mAP value of 86.9%, which is 8.7% higher than the original model. Compared with traditional steel surface defect detection methods, the proposed algorithm can detect the types and locations of steel surface defects more accurately.