Key Technologies And Realization Of PCB Opto-electronic Holechecker

With the development of science and technology, the market demand for printed circuit boards has been rising rapidly. The processing quality of PCB has been continuously upgraded and the precision and speed requirements of its online testing are also kept rising. Improving the ex-factory qualified rate of products has been an important means for an enterprise to enhance its market competitiveness. Currently, the numbers of China's high-end products of PCB testing device are very small and the foreign products are very expensive, which limits the quality improvement of China's PCB industry to some extent.The PCB opto-electronic Holechecker is a highly-intelligent PCB drill-hole automatic testing device. It provides two information preparation methods for standard boards and may automatically report many defects of PCB drill holes. The testing precision is high, at a high processing speed, which ensures thorough testing of all ex-factory PCB.In this dissertation, based on the principle of modularization, low cost and high performance, scheme design and concrete realization of the PCB opto-electronic Holechecker system are conducted. The structure of this dissertation is mainly divided into three parts, i.e.system structure design, high-speed image acquisition and software design.The system structure design part mainly includes the height adjustment system of linear array contact image sensor (CIS), which ensures clear imaging of PCB of different thicknesses; it solves the problems of maintaining synchronism between operating speed of conveyor belt and image acquisition so that acquired images will not distort at different conveyance speeds.High-precision testing requirements make the acquired images data quite large. The CameraLink is used to achieve high-speed transmission of ultra-large images. Aiming at the difference between video signals of different segments of CIS, 64-stage gain control of A/D9822 chip is adopted, to automatically set up gain parameters for different segments. The methods mentioned above have significantly improved the quality of acquired images and also facilitate the subsequent binarization processing of images.Two methods of preparing standard board information are proposed: one is to adopt the preparation method of combining drill hole documents and scanning; the other is to prepare through scanning a qualified PCB board. The algorithm proposed in this dissertation may correct the seldom-missed defective round holes.The biggest challenge to the algorithm is the real-time processing requirements of ultra-large images. Algorithms of approximate registration and accurate registration of testing PCBs and standard PCBs are proposed to ensure the precision of image registration. A large number of image processing algorithms are optimized in this paper to raise the testing speed. The adaptive scale corner detection method based on edge direction angle is proposed. 4 contour corner relationships of PCB are utilized to compress the parameter space and significantly reduce the computation load of straight-line fitting in Hough transform. The computation load of rapid Hough transform circle detection is analyzed to propose the multi-scale rapid testing method. By combining the Hough transform circle detection and the least squares circle detection, the anti-noise capability of circle detection is improved and the precision of circle detection has been obtained. The algorithm proposed herein can be used to test round holes as well as heterogeneous holes.The algorithm and theoretical researches have been successfully applied to the PCB opto-electronic Holechecker and reached the technical index of the third generation PCB Holechecker, which has been promoted to the market.