摘要
利用图像处理和模式识别技术实现了一种无损、快速且性价比高的印刷电路板缺陷检测系统。简要介绍了系统工作过程和设计原理,着重阐述了系统实现过程中的图像采集、图像增强、图像分割和图像配准等关键技术。系统实现了常规的短路、断路、突起、凹陷等印刷电路板缺陷检测,基本可以满足中小PCB生产企业的检测需求。
Machine vision-based PCB defect inspection system is designed to meet a non-destructive,rapid and cost-effective requirement for printed circuit board defect detection with image processing and pattern recognition techniques. This paper firstly introduced the whole system structure and the principle of vision detection, while described the relevant key technologies used during the PCB defect inspection,such as image collection, image enhancement, image segmentation and image registration; finally implemented one set of test system with the key technologies mentioned. The experimental results show that the defect of PCB such as short circuit, open circuit, protuberance and concavity can be effectively inspected, which can meet the examination need of small PCB manufacturing enterprise basically.
作者
王玉萍
郭峰林
Wang Yuping;Guo Fenglin(Wuhan Polytecnic University,Wuhan 430023,China)
出处
《科技通报》
北大核心
2017年第1期101-105,共5页
Bulletin of Science and Technology
基金
国家自然科学基金项目支持(61272278)
湖北省教育厅科学研究项目(B2013222)
关键词
印刷电路板
缺陷检测
自动光源
图像配准
printed circuit board(PCB)
defect detection
automatic light source
image registration
