基于颜色直方图的电路板表面缺陷检测  

Color histogram-based detection of defects on circuit boards surface

作　　者：仰梓淮 黄海鸿[1,2] 刘贺 刘赟 李新宇[1,2] 刘志峰[1,2] YANG Zihuai;HUANG Haihong;LIU He;LIU Yun;LI Xinyu;LIU Zhifeng(School of Mechanical Engineering,Hefei University of Technology,Hefei 230009,China;Key Laboratory of Green Design and Manufacturing of Mechanical Industry,Hefei University of Technology,Hefei 230009,China)

机构地区：[1]合肥工业大学机械工程学院,安徽合肥230009 [2]合肥工业大学机械工业绿色设计与制造重点实验室,安徽合肥230009

出　　处：《计算机集成制造系统》2024年第7期2296-2305,共10页Computer Integrated Manufacturing Systems

基　　金：国家重点研发计划资助项目(2019YFC1908002)。

摘　　要：为了提高废旧电路板的回收再利用率,针对回收电路板常见表面缺陷,在均匀分块基础上,提出了四叉树分裂颜色直方图缺陷检测方法。该方法能快速定位电路板表面缺陷,并通过支持向量机(SVM)实现缺陷分类,进而为电路板的二次利用提供质量保障。重点分析了分块大小与判断阈值对缺陷定位结果的影响,在保证检测精度的同时,检测速度相比均匀分块方法得到明显提升。与Faster-RCNN网络方法进行对比,结果表明该方法定位效果好,分类准确率平均达81%。To improve the recycling rate of waste circuit boards,aiming at the common defects on the surface of recycled circuit boards,a quadtree split color histogram detection method was proposed based on uniform blocking,the surface defects of circuit boards were located effectively and quickly,and the defect classification was realized by Support vector machine(SVM),thus providing quality assurance for the secondary utilization of circuit boards.The influence of block size and judgment threshold on defect location results was analyzed,and the detection speed was significantly improved compared with the uniform block method while ensuring the detection accuracy.Compared with the Faster-RCNN method,the result showed that the proposed method had a good localization effect and the average classification accuracy was 81%.

关 键 词：机器视觉 二次利用电路板 表面缺陷检测 颜色直方图 四叉树分裂 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]