YOLO-MBBi: PCB Surface Defect Detection Method Based on Enhanced YOLOv5

Printed circuit boards (PCBs) are extensively used to assemble electronic equipment. Currently, PCBs an integral part of almost all products. However, various surface defects can still occur during mass production. An enhanced YOLOv5s network named YOLO-MBBi is proposed detect on address the shortcomings existing PCB defect detection methods, such as their low accuracy and poor real-time performance. uses MBConv (mobile inverted residual bottleneck block) modules, CBAM attention, BiFPN, depth-wise convolutions substitute layers in replace CIoU loss function with SIoU training. Two publicly available datasets were selected for this experiment. The experimental results showed that mAP50 recall values 95.3% 94.6%, which 3.6% 2.6% higher than those YOLOv5s, respectively, FLOPs 12.8, was much smaller YOLOv7’s 103.2. FPS value reached 48.9. Additionally, after using another dataset, metrics also achieved satisfactory met needs industrial