Application Research Of Image Processing Technology In FOG Laminating Process

As one of the core technologies of the processing and producting of touch screen module, LCD module and other products, the FOG auto laminating technology directly affects a country's electronic manufacturing level. At present, our country does not have the intellectual property rights of the automatic FOG laminating machine, the domestic market is occupied by foreign equipment manufacturers. So, it is very important to study the FOG automatic laminating machine.Vision system is an important component of FOG laminating machine. Based on the background of the elements' location and alignment in FOG laminating, this thesis does an research on the image processing technology, including image matching to extract the regions-of-interest(ROI), sub-pixel edge detection algorithm, etc. and the motion control system and user interface of FOG laminating machine were introduced. The main contents are as follows:Firstly, positioning and alignment of laminating elements, the significance and principle of the location of the mark points are described.This thesis introduces the whole process of FOG laminating process, the overall idea of marking point positioning, including template matching extraction of ROI, sub-pixel edge detection technology, Hough transform to extract linear determination center point, etc.Secondly, this thesis introduces three kinds of existing template matching techniques, including gray matching technology, corner matching technology and feature point matching technology. Experiments and analysis are carried out on all kinds of algorithms, and we obtain several image matching algorithms which are suitable for FOG laminating machine. An improved image matching algorithm is proposed, experiments show that the matching time of the improved algorithm reduces 1~3 index levels compared to the original algorithm, and the new algorithm greatly improves the matching efficiency, and has a good anti noise and a strong adaptability.In addition, the sub-pixel edge detection technology is introduced into the vision system of FOG laminating machine, and the existing three kinds of sub-pixel edge detection methods are analyzed, including the fitting method, interpolation method and the method of moments.And this thesis proposes a new sub-pixel edge detection algorithm based on sine function, the experiment results show that the proposed algorithm can effectively improve the efficiency of edge detection algorithm and it has a good robustness. And the edge lines are fitted by the Hough transform to locate the center of mark point.Finally, the motion control system and user interface of FOG laminating machine are designed.