Based on improved algorithm Canny and sub-pixel edge detection algorithm using Zernike method

Gao Lei,Lv Jingjing,Liu Guopeng,Zhang Botao
DOI: https://doi.org/10.3969/j.issn.1000-8519.2011.07.006
2011-01-01
Abstract:To satisfy the stringent requirement of measuring accuracy of fast charge coupled device(CCD) measurement system, a novel method is proposed based on improved algorithm Canny and Zernike moments of the subpixel image edge detection methods. The algorithm to using improved Canny operators to achieve coarse edge point positioning firstly, and the pixel coordinates of edge points and gradient direction are identified at the sane time, then according to the edge point is relocated with sub-pixel accuracy by means of Zernike moment operator base on the edge point vectors and the reference threshold value, finally, the sub-pixel edge detection of the image is attained. Edge detection experiments of simulation and natural images were carried out. Compared with the Zernike moment, LOG-Zernike moment and Sobel-Zernike moment algorithms, experiment results show that the proposed method has better performances in positioning accuracy, noise immunity and detection speed.
What problem does this paper attempt to address?