Research On Image Fix Algorithm For Solder Paste Aoi

With the quick development of Electronics, traditional AOIs can't meet the requirements of detection speed and accuracy when the circuit graph in PCB becomes more and more accurate and components tinier and tinier, for which the arise of the Imagemodule Array Architecture is a good news. However, the image deviation is more complicated and is harder to fix in the new architecture. Thus, a good image fix system is needed badly to make sure AOIs work correctly.Two kinds of image deviation exist in the new architecture, which are color distortion and geometrical declination. Color distortion is caused by the defects of the image system which is basically made up of a light source, a lens and a light sensor, and every single part may have defects. At the same time, geometrical declination is caused by the mechanical deviation. For example, the imagemodule array and delivery system are two main mechanical causes.In order to fix the AOI image deviation, a image fix system is implemented according to the various types of image deviation in this thesis. Different solutions are designed and applied to the AOI system program. And of course, they are fully tested. To fix the color distortion, we calculate color increment rates using the standard RGB board, then decrease these rates in real-time images. Four-point fix method, dynamic two-point accuracy makeup are solutions to geometrical declination.After numerous experiments, it can be concluded that this image fix system can effectively correct various kinds of image deviation of AOI as expected, which helps AOIs a lot to run correctly and efficiently. It is reasonable and practical to design ideas and methods showed in this thesis, and these are valuable for developing similar systems.