A System For Automatic Inspection Of Printed Circuit Board Defects Based On Machine Vision

Printed circuit board(PCB)defect inspection is an indispensable step in the production process of electronic products.Automatic Optical Inspection(AOI)is the core technology for PCB inspection in automated production lines.However,the current high-end AOI equipment is mainly monopolized by foreign companies and the price is high;most small-scale electronics manufacturers still adopt traditional manual visual inspection methods for PCB inspection.The manual visual inspection work has high intensity and is prone to missed detection and false detection,and the efficiency is not high.For small and medium enterprises to develop a low-cost,high-efficiency automatic PCB defect detection system has become pressing needs.Therefore,the project carried out the design and research of PCB defect detection system based on machine vision.The completed work mainly includes:(1)Study the hardware composition of PCB automatic optical detection system,select and determine the hardware of each major module according to the needs of the company,including industrial cameras,industrial lenses,light sources,etc.;(2)Study the positioning algorithm of PCB board,and study the mutual information location algorithm based on Powell and the SURF feature point matching location algorithm.analyze its advantages and disadvantages,and finally select the SURF algorithm,and propose the SURF combined with distance screening(modified SURF,MSURF)Location Algorithm.The method uses the known distance prior as a constraint to accelerate the matching search of the feature points.Compared with the RANSAC based SURF feature point matching location algorithm.MSURF improves the positioning speed by 40%without affecting the positioning accuracy of the system.;(3)Study the component positioning method on the PCB board,and finally locate it by means of a combination of camera calibration and component information;(4)Based on the testing requirements given by small and medium-sized production PCB companies,the corresponding defect detection algoritthms for component missing,reversed,and wrong parts are studied.The defect detection algorithm designed can realize the detection of missing parts,anti-paste,and wrong parts of components;(5)Comprehensive Task Study A PCB defect inspection system capable of performing defect alarms,generating defect position status reports,and storing defect information files was completed.The research work on the subject and the PCB defect detection system designed have important significance and promotion role in reducing the production cost of the enterprise and improving the efficiency of the enterprise.