PCB surface defect detection algorithm based on improved YOLOv8
Aiming at the problem of low accuracy of surface defect recognition of printed circuit boards(PCBs),a PCB surface defect detection algorithm based on improved YOLOv8 is proposed.First,the improved algorithm adds the EMA attention mecha-nism module to the YOLOv8 backbone network,and the EMA module recalibrates the channel weights of parallel branches by en-coding global information to enhance the detection of PCB surface defects.Secondly,the model convergence speed is accelerated by introducing the Inner-IoU loss function to enhance the learning capability.Experimental tests are conducted on the PCB defect dataset,and the results show that the improved algorithm improves the recall rate and average detection precision by 8.1 and 2.9 percentage points,respectively,compared with the original YOLOv8 model.
万方数据
