Summary: | 碩士 === 國立臺灣科技大學 === 電子工程系 === 88 === Printed circuit boards (PCBs) assembly is a promising industry in the twenty century, especially in Taiwan. More and more financial groups invest in this business. Although the PCBs have been automated mounted in various processes, it still requires the human operators to perform the visual inspection task. The task is very consuming time, tedious, and fatiguing on the human inspectors. Thus, the application of automated inspection through machine vision has gained growing important.
We devised and developed an automatic optical inspection system for PCBs in the thesis, and gave an example of inspecting defects of SMD components. This thesis presented an inspection method based on programmable three-layer of ring-like LED illumination system and “ROI inspection window”. This method utilized the special illumination system to solve problem in inspection three-dimension objects, and employed the ROI method with databases of components’ PAD to search position of component easily. Finally, we found out missing component, misalignment, solder joint defects on PCBs by techniques of machine vision and neural network in the image of ROI.
|