A Fast Tool Edge Detection Method Based on Zernike Moments Algorithm
Liu Zixin, Liu Jinyue, Yang Li, Weng Liming · IOP Conference Series Materials Science and Engineering · 2018
due to the fast and high-precision edge detection requirements of tools in projection measurement, a fast sub-pixel edge detection algorithm for tool images is proposed. The algorithm first uses the Ostu threshold to perform the background segmentation of the tool image, then uses boundary tracking to obtain the boundary points of the image, and designs the corresponding correction operator to obtain the edge information of the tool; then according to the edge gray model, the Zernike moments are used to calculate the gradient direction of the edge points, and the two-dimensional model of the edge gray is converted into the one-dimensional model, and the sub-pixel edge of the tool image can be accurately obtained by calculating the gray distribution by spline interpolation. Experimental results show that this method has better noise immunity and faster detection speed.