摘要
设计了一种针对微小零件闭气塞的表面缺陷自动检测系统。该系统采用自动控制技术实现闭气塞的上料、排序、搬运、定位和正废品分离,采用图像采集和处理技术实现缺陷的自动识别与分类。选择快速图像匹配算法快速定位目标零件在图像中的位置,把目标零件和无缺陷的标准模板图像减影获得缺陷图像,以缺陷的几何特征和统计特征获得缺陷的类型。实验结果表明,该方法可以实现闭气塞表面缺陷的自动检测检测与识别,识别分辨率可以达到5.6μm。
An automatic surface defect detection system is proposed for small component gas lock valve. In this system, the automatic control technology is used to realize loading, sorting, carrying, orientation, separation for the part, while the image grabbing and processing technology are used to realize detection and automatic classification of part surface defect. Fast image matching is used for locating the gas lock valve in the grabbing image, then the part image is subtracted from standard template image in order to obtain the defect image. At last the defect classification is obtained by the defect＇s geometrical and statistical characteristic. The experiment results show that the method can discern surface defect for gas lock valve,and the resolution can reach 5.6μm.
出处
《国外电子测量技术》
2011年第4期13-16,共4页
Foreign Electronic Measurement Technology
关键词
机器视觉
闭气塞
表面缺陷
自动检测
machine vision
gas lock valve
surface defect
automatic detection
