Research On Key Technologies Of PCB Automatic Optical Inspection Machine For Appearance Defects Based On Machine Vision

With the rapid development of electronic industry,the processing accuracy of PCB is also increasing.In order to effectively reduce the rejection rate of the products,the detection of PCB appearance defect is very important in the actual industrial production.The manual visual inspection was firstly used in the PCB appearance defect detection.With the increasingly integrated highly production of printed circuit boards,manual visual inspection can't meet the requirements of reliability.For the repetitive,monotonous and strict inspection tasks,optical automatic detection is the best solution.PCB appearance optical inspector is a multi-module integrated system which integrates high uniform brightness illumination system,high-speed linear array camera image acquisition,high-precision servo drive system,corresponding image processing algorithm and software technology.The developed countries,such as the United States,Japan and Israel,are in the lead in the research and development of PCB appearance optical inspection machines,and have achieved commercialization of certain scale results.The domestic region start in this field is relatively late,and there is a big gap with the advanced level of foreign countries.Based on the research results of PCB appearance defects at home and abroad,this paper presents a set of hardware structure and a set of detection algorithm for PCB appearance defects.Firstly,the influence of camera hardware,light source type,color temperature of light source and illumination mode on PCB imaging and the basis of selection are discussed.In order to reduce the image distortion in the process of image acquisition.a coordinated motion control scheme based on multi-sensor signal fusion is proposed.Secondly,through the study of PCB defect detection algorithm,a complete set of PCB standard plate information generation algorithm,PCB automatic optical detection algorithm based on image registration,and defect classification algorithm based on the local pixel distribution characteristics of defects are proposed.Finally,the whole PCB appearance defect detection experiment is designed,and the realization idea of the software part of the PCB appearance defect inspection machine system is expounded.The feasibility of the detection algorithm is verified by the analysis of the experimental results.The results show that the proposed algorithm can accurately detect the appearance defects of PCB,realize contactless automatic optical detection,improve the efficiency of the algorithm,and basically achieve realtime detection.