|
The subject of this study is based on image processing technology, quality inspection of solder paste printing. The main application of the technology in the production line of surface mount technology (SMT, Surface Mounted Technology)-based automatic optical inspection (AOI, Automatic optic Inspection) system, in recent years, more and more types of testing equipment majority Electronic Component Manufacturers favor, its operation is simple, fast operation, supporting the use of the state-of-the-art SMT production line AOI testing equipment can not only increase productivity, but also a strict check on the quality of the products, the technology has a huge market prospect and research value. 's. Solder paste print quality inspection is the first detection SMT production line automated optical inspection process, follow-up component placement detection and chip placement testing there. In order to describe a convenient the AOI system discussed later if no special statement, refer to the for paste printing quality detection AOI system. This article first introduced the the AOI system works, AOI equipment is a blend of electric optics, physics, computer vision and image processing technology in one system. The system includes not only on paste the image processing and analysis software, further comprising a motion control device and the data management system on the analysis results save feedback. During solder paste printing quality testing before we must first locate on the printed circuit board (PCB, Printed Circuit Board) PCB positioning located a pair of diagonal reference point (Mark points), to find this on the position of the reference point coordinate changes, relocation of the PCB. Pyramid model, and obtain the reference point area where the first through the thickness of the template matching, Mark point rules Pattern it in the communicating area in the center is the center of the reference point and therefore can be precisely connected domains center find Mark the center of the point, to lay the groundwork below to get accurate solder paste connected domain. Solder paste on the PCB image rich in color and texture information, first solder paste image color and texture of the statistical analysis, modeling, segmented binary image of the solder paste position; position binary of solder paste image morphological processing information connected domain; then use the template information than on ways to detect whether there is a short circuit, multi-tin, small tin defects exist. After several parameters to adjust, reliable detection.
|