Research On Precise Positioning System Of PCB Automatic Testing Machine Based On Machine Vision

With the continuous development of PCB in the direction of fine pitch and high integration,the requirements for PCB testing technology is becoming more and more high,and the visual system has become one of the important factors that affect the PCB test.According to the requirement of the PCB automatic testing system of the visual system,the automatic testing visual system is selected as the research object and the visual system architecture is built reasonably.Meanwhile,the camera calibration method,visual localization algorithm and automatic test platform structures,as well as the system error and the compensation,are comprehensively studied.Firstly,for the requirements of the PCB automatic testing system in intelligent positioning,the camera calibration method was studied to establish the camera model under geometric distortion conditions.And then,the Bouguet sub-pixel corner extraction algorithm of the checkerboard calibration board is mainly focused on,at the same time,the camera is calibrated based on the Zhang Zhengyou's camera calibration method.The simulation and experimental results show that the camera parameters obtained after the algorithm calibration are accurate and stable,and the accuracy is high.After the camera calibration is completed,the Mark point and measurement point positioning algorithm on the PCB board are studied.For the Mark point positioning,the subpixel least square fitting algorithm is selected as the algorithm of the localization of Mark point by evaluating the positioning accuracy and positioning time between the normalized cross-correlation visual positioning algorithm and the subpixel least square fitting algorithm.For the measuring point positioning,the error model is established through analyzing the position error and size error of the circuit board.Due to the transformation between the displacement measurement value and the coordinate system of the control platform is realized through the decoupling algorithm,the measuring point positioning of the circuit board is completed.According to the requirement of PCB automatic testing system to the image quality of vision system,the camera,lens and light source of vision system are selected in order to determine the illumination mode,and the method to improve the testing precision is given through analyzing the measurement error of vision system.Based on the requirements of the PCB automatic test system for the motion control platform,the motion control board,driver,and motor are selected to build a reasonable motion control platform.After the installation error of the control platform was analyzed,the error of the system is compensated.Finally,the host computer interface and control program is designed to realize the integration between the vision system and the motion control system by choosing the Lab VIEW and MATLAB as the software system of automatic test system.