Circular Saw Blade Image Feature Extraction Based on Computer Vision
According to the image features of the circular,the image preprocessing,contour extraction,contour optimiza-tion,detection of the center of the saw blade,measurement of the diameter of the inner and outer circles,measurement of roundness error,detection of the tip and two straight lines of the tooth tip,measurement of the tip angle,and measurement algorithms of radial rake angle and radial back angle are integrated in this paper.According to the characteristics of the tooth tip of the circular saw blade image,the Hough transform of the straight line is improved,the collinear pixel points around the tooth tip at each tip are detected first,and then the two straight lines around the tooth tip were detected by the Hough transform one by one.In order to solve the long time of the detection of the conventional Hough transform of the circle,the Hough transform of the circle is improved.The total number of pixels at the edge of the circle is divided into three equal parts,and then one pixel is taken from each part of the edge point,and the three pixels that are not collinear are used to determine the center of the circle.This greatly shortens the detection time on the premise of ensuring accuracy and lays the foundation for the on-line inspection of parts.
万方数据
维普
