Research on the Method of Identifying the Collapsed Angle Surface of Fine Punching Small Gear

Abstract

Aiming at the requirement of automatic assembly of automobile rear wiper for automatic identification of the collapsed corner surface of finely-punched pinion, this paper proposes a method based on the mean value of the area of multi-toothed region of finely-punched pinion as the characteristic parameter, which can better identify the collapsed corner surface. Firstly, the industrial camera is used to obtain the pinion image, and the Halcon operator is used to fit the circle in the pre-segmented region to obtain the pinion circle position parameter, and the parameter is used as the reference to accurately segment the toothed annulus ROI, and then carry out the dynamic thresholding of the region, merging the regions, shape convex package transformation, and the conditional filtering to obtain the multiple toothed regions, and then rank the toothed regions, and the maximum top three toothed regions area mean value is the feature parameter to achieve the fine blanking pinion surface automatic identification of the collapsed corner surface. The mean value of the area of the first three tooth-shaped areas is used as the characteristic parameter to achieve the identification of the collapsed angle surface of the fine blanking pinion gear. The application on the assembly line shows that the recognition accuracy of this method reaches 99.9%, with good robustness and stable operation, which meets the production requirements, and also provides a method for the recognition of collapsed corner surfaces of other fine blanking parts.