Research And Design Of Defect Inspection System For Printed Circuit Board Based On Image Processing

The automotives parts have extreme requirements for the reliability of the products,and the motor control system is one of the highest parts of fault rate in automotives parts.It is very important to make sure the quality of the printed circuit board(PCB)used in the motor control system.In addition to the inspection of electrical parameters to ensure the reliability of PCB,it is also important to inspect the reliability of the surface of PCB and the welding board.The traditional way of human inspection is short in the checking speed and precision.Since the checking speed and precision of image processing is pretty good in modern inspection industry so the way of image processing to inspect defects of PCB is more and more important in modern industry.This paper is aimed at the defect detection of printed circuit board used in the high power driving circuit of automobile electronic control system.Firstly,the characteristics of high-power drive circuit are analyzed,and the defect types and corresponding characteristics of printed circuit boards used in high-power drive circuit are defined.The functional requirements of defect detection and defect type identification and location for bare-board conductor defect and solder joint defect are proposed.Based on the comparison of the existing PCB defect detection technology,the PCB defect detection and recognition system scheme based on image processing is determined.According to the system scheme,the basic process of defect detection based on image processing is designed,including image acquisition and pre-processing,image processing and defect detection and recognition.According to the experimental results of the algorithm,the image processing algorithm of the system is determined by selecting the best one.The standard PCB image is taken as the reference image,and the defect detection of the PCB image is carried out.At the same time,the recognition algorithm of different defects is designed,which realizes the defect detection system of PCB.At the same time,a human-computer software is designed for the system.The test results show that the system can detect and identify the defects of different conductors and solder joints of PCB accurately to ensure the feasibility of the system;and in the process of batch experiments,the detection accuracy of conductor defects is as high as 92%,and the average recognition accuracy of different conductor defects is 92%.The detection accuracy of solder joints defect is 80%,and the average recognition accuracy of different solder joint defects is 86.67%.The result of the system is higher than that of other algorithms.It proves the reliability and superiority of the system and realizes high efficiency,high precision and diversity defect detection.