Screen Printing Electronic Printed Circuit Detection Technology

With the development of electronic products, electronic screen printing applications have been very extensive. In daily life, the printed electronic screen has been used in smart phones, computers and readers shield. High precision electronic circuitry has been printed on the back panel of printed electronic screen. Due to the interference of the printing environment, printed circuit process is prone to short circuit, burrs and other defects, but published research specifically for the detection of such printed electronic screen with printing problem is minimal. This article focuses on the technology involved in the course of the detection of printed circuit on the printing electronic screen.First, since the back surface of the printed circuit electronic screen plated with a layer of bluish film, so that a color image captured with uneven component at R, G, B three-channel, and the R channel component is very small. If the image is gray-scale processing, as the R channel with a large proportion of component, causing the divided image affected. To solve this problem, this article propose a improved segmentation method based on Otsu, reducing the impact on the image defects. The improved method separates the color images into three channels directly, and then using Otsu segmentation. To getting the image segmentation we should vote to determine the pixel by pixel with obtained three sites images and three segmentation threshold, and ultimately getting the whole image segmentation. Due this method eliminates the image gradation process, and thus make the greatest efforts to reduce the impact of graying caused.Secondly, when purifying the feature point matching, this paper presents a method based on the spatial distance purification. Based on different image rotation angle is small, the spatial position of the same feature point in different images should be very close. So different images should be placed in the same coordinate, then calculate the spatial distance between the same pair of feature points, when the distance is greater than a predetermined threshold is considered mismatch. The method can not only be used for matching individual purification can also be combined with RANSAC method and ratio purification method. Experimental results show that the error matched can be culling effectively by this method alone, it can improve the speed and accuracy when combined with RANSAC.In addition, according to the characteristics of the printed circuit printed electronic screen, after experiments this paper choosing the method by controlling the size of the communication domain for image filtering, experiments show that this method can effectively filter out all the noise block.Finally, under VS2010 software environment, the algorithms proposed in this paper and the whole process of printing electronic screen printed circuit were implemented. The results show that the proposed method can effectively defect image obtained.