An automatic inspection system for the coating quality of the edge of mirror elements

This study aims to inspect the common defects of the edge of the mirror elements, including scratches, impurities, and edge serration, which occur in the production of mirror elements, where a mirror surface was illuminated in different directions in the experiment. The imaging system was used to inspect the defect position, enhance image contrast, and judge whether the defect exceeds the permissible range. With a novel algorithm, the image contour was analyzed by edge detection, and combined with edge evaluating parameters to improve the quality of the coating edge. According to the experimental results, the proposed algorithm has an average recognition rate of 96% and the recognition speed on a mirror element sized 10 × 7 cm is only 0.1 s.