基于YOLOv5的PCB缺陷检测研究

Research on PCB Defect Detection Based on YOLOv5

印刷电路板作为电子设备的核心部件,其质量决定了电子设备的功能运转,因此高效、精准的检测PCB缺陷十分重要。本文将使用YOLOv5目标检测算法,实现PCB缺陷识别检测。通过实验得出在6种PCB缺陷类型中对于漏孔类型缺陷检测精度达到98.9%,所有类别mAP可达95.4%。实验结果表明,YOLOv5算法可以满足工业生产中缺陷检测的需要。

As the core component of electronic equipment,the quality of printed circuit board determines the functional operation of electronic equipment,so it is very important to detect PCB defects efficiently and accurately.This paper will use YOLOv5 target detection algorithm to realize PCB defect recognition and detection.Through experiments,it is concluded that the detection accuracy of six PCB defect types for leak type defects reaches 98.9%,and the mAP of all types reaches 95.4%.The experimental results show that,YOLOv5 algorithm can meet the needs of defect detection in industrial production.