基于计算机图像处理的电路印刷板缺陷检测

Research on Defect Detecting Method of Printed Circuit Borad

针对人工目检等传统PCB检测方法高成本、低效率问题,通过提出一种基于计算机图像处理的PCB板计算机视觉混合检测方法,完成了实现快速稳定可靠的PCB缺陷智能在线检测的目的;建立了系统的软硬件结构,研究出结合图像分模块处理,改进的极小值分割、快速模板匹配、图像对比、形态后处理等措施的混合智能检测算法;实验结果证明,该方法解决了海量数据与实时性处理的矛盾以及噪声干扰等问题,能准确检测PCB常规缺陷,并保证了自动在线检测的实时性、可靠性和低成本。

The purpose of PCB automatic online inspection is realized by providing a kind of design of PCB defect inspection based on image-processing. The hardware and software structures of system was built and hybrid inspection algorithm was developed which integrated some efficient measures such as image processing in block, Threshold Segmentation Of Minimum Value. Fast Template Matching and Mor-phological processing. The result of experiment indicated that this method resolved the conflict between datas and time, as well as the disturb of noise, detected the general defect of PCB, and realized the automatic online detecting fast. safely and economically.