Research On LED Paste Quality Inspection System Of Flexible Printed Circuit Board

Flexible printed circuit board(FPC)is a kind of thin,light and foldable printed circuit board,which is more and more widely used in the field of electronics.The FPC studied in this paper is mainly used in computer keyboards,the quality of patch-type LEDs on FPC will directly affect whether an FPC can enter market.Due to the special nature of FPC materials,as well as the small size,large deformation and variety of LEDs,manual visual inspection is very difficult,and PCB detection methods cannot be applied to FPC quality inspection.In this paper,a set of on-line inspection method of LED stickiness quality based on machine vision is designed to meet the requirements of existing productivity for the problem of leakage and sticking defects in the production process of flexible printed circuit board LEDs.The main contents of this paper are as follows:(1)According to the position distribution of the LED on the FPC and the shape characteristics of the LED,a suitable hardware system is designed to obtain the image,including camera,lens,light source selection and spatial position design,as well as design the camera's motion path and trigger mode.(2)The software system is designed according to the overall functional requirements,including designing human-computer interaction interface,camera calling module,detection algorithm module,multi-thread architecture and software communication module.(3)Implement LED paste quality detection algorithm.Firstly,In order to minimize the impact of interference areas on judging LED quality,a large of image is firstly pre-treated,including gray correction,binarization,morphological manipulation,and geometric features is reduced the influence of interference areas.Secondly,the binary pad template matching is used to realize the coarse positioning of the pad area,and a single pixel pad profile is obtained as a positioning template by a template parallel traversal idea,and then the fine positioning of the pad is achieved by using an asynchronous scaling matching method.Then the Gaussian bilateral filtering is used to remove the noise in image and a kind of method of wick region roundness is calculated by region projection is proposed.According to roundness patch of different contrast adaptive parameter canny edge detection for getting wick edge,the wick edge is fitted to obtain wick coordinates.Finally,the defect quality of LED is judged according to the deviation between the welding plate coordinate and the wick coordinate,and the geometric defect of the image is converted into the numerical judgment of the coordinate.Finally,the LED hardware imaging system was built,the online detection method for defects of 2mm×3mm LED was designed.The average time for detecting a large image of 8192×70000 pixels was 1538.2ms,leak detection accuracy is 100% and the accuracy of stickiness detection is 97%.The inspection time and accuracy are in line with the requirements of the industrial site.