A Novel Rapid Approach to Image Edge Detection

ZHANG Ke,SUN Xiao-wei,LI Yan-jun,LIU Yan
DOI: https://doi.org/10.3969/j.issn.1673-9728.2006.01.094
2006-01-01
Abstract:Edge detection is the base of feature extraction,analysis and comprehension of image,the quality of which determines the results of subsequent processing.Thus it is an effort goal for people to find a kind of method that is rapid,precisely locates true edges and excludes false edges.This paper discusses and tests a new rapid approach to edge detection based on standard deviation.We compute the standard deviation of the pixel's neighboring region to determine the edge alternation,then we detect the edge pixel.The experimental results show that this approach presented by this paper can obtain more precise edge pixels than some traditional edge detection algorithms,such as sobel and Gauss-Laplace.So the method is a simple and effective edge detection algorithm.
What problem does this paper attempt to address?