Hole Defect Detection For High-Density Flexible Integrated Circuit Packaging Substrates

High-density flexible integrated circuit packaging substrates are widely used in various electronic products due to their excellent characteristics,such as light weight,thin thickness,high flexibility,high wiring density and so on.However,the production process is numerous and the manufacturing process is complicated.So the defect detection in the production process is particularly important.The round hole is an important part of the connection between the flexible substrate and the external circuit,and its appearance quality plays a key role in packaging and assembly.Regarding the hole defect detection,traditional manual visual inspection methods cannot meet the production needs of enterprises.Existing visual detection methods are usually designed for detecting the specific round hole.However,there are many types of round hole components in flexible substrates,and their defect manifestations are random and uncertain.Therefore,the defect detection for single type of round hole is difficult to apply to actual industrial production.In response to the above problems,this paper designs a general hole defect detection method for high-density flexible integrated circuit packaging substrates.The main processes of the method are image preprocessing,target contour extraction,abnormal hole recognition,and defect decision.In-depth research on key processes are as follows:(1)A target contour extraction algorithm based on geometric active contour model is proposed,which transforms the extraction of target contour into curve evolution,and uses an improved variational level set method to solve it.Finally,the curve is discretized to obtain a pixel-level target contour.(2)A sub-pixel target contour extraction algorithm based on Lagrange polynomial is proposed.The algorithm uses Lagrange polynomial as the interpolation function.According to the target contour extraction results in this paper and the gradient information of the image,the sub-pixel contours near the pixels can be extracted,thereby improving the detection accuracy.(3)Based on the result of target contour extraction,a fast feature construction method is proposed,and an abnormal hole fuzzy recognition algorithm using parameter estimation is proposed according to the constructed feature vector.It uses the maximum likelihood method to estimate the parameters of the membership functions.The algorithm can quickly identify abnormal holes on the flexible substrate.(4)According to the hole defect detection standard,the parameter calculation method is designed.Using contour feature vectors,applying cubic spline curve fitting,color space transformation and other theories to calculate the target position,minimum radius,deformation degree,completeness,connectivity,hole vacancy size and other parameters.In this way,hole defects on abnormal hole targets and circular parts can be detected.This paper solves the problem of automatic detection of hole defects in the production process of high-density flexible packaging substrates,which has certain theoretical significance and engineering practical value.