Texture Segmentation of Natural Images Based on Active Contour Model

Abstract

To segment complex texture natural environment images; the first, the texture features of natural images should be analysed and the texture features should be extracted; The second, texture images segmengtation can be achieved by using Mumford-Shah active contour model, this segmentation model can better process fuzzy, default boundary, and this model can be solved by level set method. This method can express well complex texture signal features of natural images. Through making texture segmentation experiment for standard texture synthesis image and natural environmental image, its results show that the texture segmentation based on Mumford-Shah active contour model can segment natural images.