Applications Of Automated Optical Inspection System In SMT

During the early stage of PCB production, the testing is mainly done by the artificial detect with the power. With the development of electronic technology, PCB wiring density increases continuously. The difficulty for the artificial detect also increases with higher rate of miscarriage of justice and more harm to the detector's health. The programming of electricity detection becomes more cumbersome and expensive. Besides, it is not able to detect certain types. Therefore, Studying on AOI is has important significance not only on science but also on practical.Automated optical inspection system in the Application of SMT is composed of three parts, which are image capture, image pre-processing and feature extraction, and feature determination. This dissertation deals with these key problems in hardware and software of AOI, and presented some new ideas, approaches and come to some valuable results:Firstly, based on the research and analysis of the current AOI technology and the present advantages and disadvantages of different motion controller, the thesis proposes design of embedded motion controller with ARM+DSP as the main processing platform..Secondly, aimed at the full use of available resources both at home and abroad and the reduction of developing costs, the thesis adopts edge detection algorithm in OpenCV and optimizes the algorithm, making it more convenient for appliance.Thirdly, the thesis makes a deep research of another main part of AOI which obtain image information of the detected object, including the interface between the camera and CPU, and the camera driver.Finally, the defects often occurred in SMT soldering are dealt with successfully by AOI technology.The experimental result shows that embedded controller as the main processing platform combined with edge detection algorithm in OpenCV is particularly suitable for the environment of industrial production. The edge detection algorithm in OpenCV and its optimization can make the world's leading digital image processing algorithms apply timely to practical production.