Improved Segmentation Technique Based on Graph Theory
Jing Zhang, Shanshan Zhao, Tianchi Zhang, Geng Chen, Jing Yang · 2018
A method based on graph theory is proposed, which adjusts the parameters by using a method of chi square with Gauss weighted distance and adaptive parameter adjustment of shared neighbor weighted, in this paper. The pixels are arranged with the vector form of local neighborhood information, considering the different effects of each pixel in the neighborhood of the central pixel, the weight should also be different, the Gauss function will be used chi square distance combinations, to avoid over segmentation or under segmentation. And by integrating the nearest neighbor weighted adaptive method, each pixel is automatically given a scale parameter to reduce the need to adjust the parameters. Gauss weighted local neighborhood information is introduced in this paper, to construct the similarity matrix directly on the original image.