无线电工程2024，Vol.54Issue(7)：1614-1621,8.DOI:10.3969/j.issn.1003-3106.2024.07.003
基于改进YOLOv8的PCB焊点语义分割方法
Semantic Segmentation Method for PCB Solder Joint Based on Improved YOLOv8
卢子册 1刘小芳 2王德伟2
作者信息
- 1. 四川轻化工大学自动化与信息工程学院,四川宜宾 644002
- 2. 四川轻化工大学计算机科学与工程学院,四川宜宾 644002
- 折叠
摘要
Abstract
Aiming at the problem of false detection and missing detection of Printed Circuit Board(PCB)solder joints with similar gray values in the process of detection and segmentation,a semantic segmentation model for PCB solder based on improved YOLOv8 is proposed.Firstly,Coordinate Attention(CA)mechanism is introduced into the backbone network to accurately locate the spatial position of solder joints and to improve the ability of the model to capture the spatial information of solder joints.Secondly,Bi-directional Feature Pyramid Network(BiFPN)is used to replace the Path Aggregation Network(PANet)feature pyramid to better capture the boundary information of the target,and add a segmentation layer on the basis of the original.Finally,the EIoU loss function is introduced to provide finer evaluation results and improve generalization.Comparative experiments show that the mean Pixel Accuracy(mPA),mean Intersection over Union(mIoU)and Frames Per Second(FPS)of the proposed algorithm reach 90.37％,83.76％and 43 FPS,which realizes more accurate solder joint segmentation of PCB board.关键词
YOLOv8/语义分割/坐标注意力机制/双向特征金字塔网络/EIoU损失函数Key words
YOLOv8/semantic segmentation/CA mechanism/BiFPN/EIoU loss function分类
信息技术与安全科学引用本文复制引用
卢子册,刘小芳,王德伟..基于改进YOLOv8的PCB焊点语义分割方法[J].无线电工程,2024,54(7):1614-1621,8.
