A Welding Defect Detection Algorithm Based on Deep Learning

Abstract

In order to meet the needs of process inspection technology for industrial equipment, image recognition technology based on deep learning has shown great potential in the field of welding defects. In this paper, an improved YOLOv8 algorithm is proposed to improve the welding defect identification ability of the workpiece. Through experimental verification on selected data sets in kaggle, this study evaluates the detection performance of YOLOv8 improved algorithm that integrates SCConv in C2f module at Backbone level. The experimental results show that the improved YOLOv8 has improved the accuracy of welding defect detection compared with the traditional version, and has certain application potential.