Research And Application Of Optical Inspection Technology For PCB Bare Board Defects

With the rapid development of electronic information technology,PCB is becoming more and more high density and multilayer,traditional detection methods cannot meet the requirements of the PCB board industrial inspection,AOI detection with its advantages of high efficiency,high precision and no damage,is becoming the development direction of PCB defect detection.In order to detect defects in short circuit,open circuit,sag,bump,cavity and remaining copper,we need to design corresponding AOI defect detection system to detect defects in PCB bare board,and meet certain accuracy.Firstly,the overall design of AOI detection system is carried out,which is divided into lighting unit,image acquisition unit and image processing unit.According to the size and precision requirements,select the appropriate camera and lens,select the appropriate light and lighting mode.Secondly,the research of image preprocessing algorithm is completed.By comparing different gray scale algorithms,the weighted method is selected to make the color image grayscale.By comparing the commonly used enhancement algorithms,the histogram enhancement is selected to improve the contrast of the image.By judging the source and type of the noise,the Butterworth low pass filter is selected for smoothing the noise.The commonly used threshold segmentation algorithm is studied.The 2-D maximum variance method is applied to thresh the segmentation.It reduces the amount of computation by dimension reduction,and combines it with genetic algorithm to accelerate the optimization.The experiment proves that its operation speed is increased by about 50%.Thirdly,the research of image registration algorithm is completed,and feature circle matching is selected to complete the image registration.Compared to different algorithms,the Canny algoritihm is selected for edge extraction.Morphological treatment was used to improve the quality of the edges.Hough transform circle detection is used to detect the parameters of the characteristic circle,and the computation speed is accelerated by applying the constraint conditions.It is combined with the least square method to improve the detection precision.The experimental results show that the relative error of the characteristic circle is reduced and the accuracy is improved.Finally,the reference comparison method is used to detect the defects.This paper compares the connected domain method with the contour method,evaluates the accuracy and speed of operation from two aspects,and chooses the contour method to identify the defects.MATLAB is used as the development platform to integrate and test the software system.Experiments show that the system meets the requirements of detection.