一种改进YOLOv5模型的PCB板缺陷检测方法
A method for detecting surface defects of solar cells based on improved YOLOv5
陈怡菲 1汪繁荣 1鲁东林 1刘逸凡2
作者信息
- 1. 湖北工业大学电气与电子工程学院,武汉 430070
- 2. 华中科技大学武汉国家光电研究中心,武汉 430072
- 折叠
摘要
针对传统PCB板检测方法中检测效率不高,检测精度较低等缺点,提出了一种改进YOLOv5模型的PCB板缺陷检测方法.为提升小目标缺陷检测精度,构建了基于BiFPN的网络连接方式,更加充分地利用了不同尺度的特征信息;为了更好地捕捉目标缺陷的位置,引入了 Coordinate Attention注意力机制,使模型的定位和目标捕捉更加精准.实验结果表明,较原始的YOLOv5模型,所提出的针对PCB板表面缺陷检测方法的均值平均精度提高了 3.2％.
Abstract
Aiming at the shortcomings of traditional PCB board inspection methods,such as low detection efficien-cy and low detection accuracy,a PCB board defect detection method with improved YOLOv5 model was proposed.In order to improve the precision of small target defect detection,the BiFPN based network connection method is con-structed,which makes full use of the feature information of different scales.In order to better capture the position of target defects,we introduced Coordinate Attention mechanism to make model positioning and target capture more accu-rate.The experimental results show that compared with the original YOLOv5 model,the mean average accuracy of the proposed method for detecting PCB surface defects is improved by 3.2％.
关键词
PCB板/YOLOv5/缺陷检测/深度学习/BiFPNKey words
Solar cell/YOLOv5/defect detection/deep learning/BiFPN引用本文复制引用
出版年
2024
NETL
NSTL
万方数据