Intelligent Inspection System For Surface Defects Of High-Precision Flexible Integrated Circuit Substrates

Flexible Integrated Circuit Substrates(FICS)are developing in the direction of high precision,as an essential carrier for connecting chips and electronic components.Automatic Visual Inspection(AVI)technology is an inevitable trend for the IC packaging and testing industry to develop in the future.Compared with manual inspection,the use of AVI technology to detect the surface defects of the FICS has incomparable advantage in the production speed and labor cost.In this regard,aiming at the application and research of AVI technology in highprecision FICS surface defects detection,the following work is mainly completed:(1)In response to the inspection requirements of the production process for surface defects detection of FICS,the key hardware components of the optical inspection system were compared and selected,and a set of dual vision imaging systems was built.(2)This paper designed a Graphical User Interface(GUI),simple and friendly.Furthermore,using design pattern technology to construct the system's software framework to make it rich in expandability,using multi-threading and thread pool technology to optimize the running process.the efficiency of the dual vision image system is greatly improved.(3)Aiming at the defocusing problem of the metallographic microscope during the surface defects inspection process,this paper proposes a local ternary pattern texture operator based on the median value to describe the image texture and constructs an image quality assessment function.To achieve fast autofocus,an optical defocus model of the image quality assessment function value and the position of the motor was established,based on the derivation of the optical imaging model.Compared with other algorithms,the algorithm proposed has better performance in Root Mean Square Error(RMSE),Linear Correlation Coefficient(LCC),and Average Running Time(ART).(4)Aiming at the surface defect detection of FICS with complex backgrounds,this paper uses a Fuzzy C-means clustering algorithm based on local information for image segmentation,and combines the Scharr operator and the Sobel operator to enhance the weak edge information.On this basis,the multi-dimensional features of contour shape,contour internal region,and contour external space relationship were fused.Then using the minimum risk Bayesian classifier model to quickly determine the surface defects.The experimental results show that the fusion feature classifier has excellent performance and a high recognition rate.In this paper,an intelligent inspection instrument combining software and hardware is designed for high-precision FICS surface defects detection,which realized the key technology of automatic focusing with metallographic microscopes and the rapid detection algorithm for surface defects.It has certain theoretical significance and practical engineering application value.