Automatic level bar inspection system based on mathematical morphology

YU Ting, ZHU Shanan

(College of Electrical Engineering, Zhejiang University, Hangzhou 310027, China)

Abstract: Aiming at the automatic inspection of the level bar, the edge detecting and contour tracking algorithm in the field of machine vision were studied, and an automatic detection method for the lever bar based on mathematical morphology and contour tracking was proposed. In order to improve the efficiency of the median filter algorithm, a fast median filter algorithm and multithreading tool were used to deal with the image of the bubble in the level bar taken by the industrial camera. In order to get the edge information of the bubble, the fuzzy minimization morphological gradient was introduced to deal with the image of the bubble. Aiming at this application, grayscale profile of the image was analyzed toposition theparallel reference lines, and the target neighborhood point method is used to search the position of bubble around vertices, and calculate the deviation of the bubble, so as to achieve highefficiency and highaccuracy deviation detection of the level bar. The result indicates that the proposed method can greatly improve the calculating speed while ensuring accurate results at the same time, and can well adapt to the light changes.

Key words: image processing; level bar; mathematical morphology; contour tracking