Development Of Large-size PCB Alignment System Based On Multi-camera

Target alignment technology is one of the key technologies in PCB manufacturing process.Its efficiency and precision directly affect the efficiency of PCB production and the quality of finished products.This paper studies and develops large-scale PCB alignment systems based on machine vision technology.Compared with the traditional low-resolution small-field alignment method,the high-resolution large-field alignment method of the system has the advantages of large target alignment range,wide applicatio n range,and more flexible and quicker product changeover.The method of template matching in the system can effectively reduce the complexity of the operation and improve the applicability of the alignment system.The edge detection method can optimize th e positioning result to improve the accuracy of PCB alignment.(1)According to the specific alignment requirements in the PCB manufacturing process,combined with the working principle of machine vision components,the implementation scheme of the alignme nt system is designed.According to the specific alignment requirements in the PCB manufacturing process,combined with the working principle of machine vision components,this paper designs the implementation scheme of the alignment system.By calibrating the industrial camera to obtain the internal and external parameters of the camera,the distortion of the PCB image captured by the camera is corrected and the camera coordinate system is calibrated to obtain the mutual conversion relationship between the camera coordinate system and the external device(laser)coordinate system.Get a unified world coordinate.Compared with the calibration method of the traditional calibration camera and camera coordinate relationship,the calibration method adopted by the system has the advantages that the positions of the cameras are independent,the precision is higher,and the implementation is faster and simpler.On this basis,multiple cameras can be added to meet the needs of a variety of application scenarios.Three commonly used circular positioning methods—Hough transform circular positioning method,least squares fitting circular positioning method and template matching circular positioning method are studied and analyzed.Combined with the actual production requirements and data analysis,the geometric profile-based template matching method is finally selected for circular positioning.Finally,the edge detection method is used to optimize the center positioning,which improves the accuracy of the alignment.The registration system was developed using the open source image processing library OpenCV.The system software carried out long-term continuous production test in the project site,which verified the reliability and stability of the system and the accuracy and efficiency of the algorithm.