摘要
针对当前FPC(柔性电路板)缺陷检测中人工目检效率低的问题,基于机器视觉技术设计了一套实时检测系统。首先搭建了硬件系统,然后对FPC的4种表面缺陷特征进行了研究,基于Halcon设计了相应的缺陷检测算法,提出了通过模板匹配提取ROI的方法,以及运用图像自乘与高斯线检测来提取折痕,最后基于MFC开发了缺陷实时检测系统。实验结果显示,设计的系统检测准确率可达90%以上,且每片FPC检测时间只需0.2 s。
Aiming at the low efficiency of manual visual inspection in FPC(flexible printed circuit)defect detection,a real-time detection system based on machine vision technology was designed.Firstly,the hardware system was built up.Four FPC sur-face defects features was restearched,the corresponding algorithm was designed based on Halcon,and the method of extracting ROI by template matching and extracting crease by image self-multiplication and Gaussian line-detection was proposed.Finally,the real-time detection system based on MFC was developed.The experimental results show that the system can achieve more than 90%accuracy,and each FPC detection time costs only 0.2 s.
作者
眭石军
廖平
SUI Shi-jun;LIAO Ping(College of Mechanical and Electrical Engineering,Central South University,Changsha 410083,China)
出处
《仪表技术与传感器》
CSCD
北大核心
2020年第9期64-68,共5页
Instrument Technique and Sensor
基金
国家自然科学基金资助项目(51275535)。
关键词
机器视觉
FPC
缺陷检测
模板匹配
图像自乘
高斯线检测
machine vision
FPC
defect detection
template matching
image self-multiplication
Gaussian line-detection
