非参数邻域统计的无监督纹理分割方法

Unsupervised Texture Segmentation Method with Nonparametric Neighborhood Statistics

介绍了一个新颖的无监督分割方法,这种方法依赖于一个通用的图像邻域的非参数统计模型,直接建模图像邻域,不用建立中间特征.它不是针对某种特定纹理,而是通用在各种纹理上.文章通过静态随机域和非参数的高阶统计模型探讨了图像纹理的基本描述.文章中提到了适合各种纹理的通用的公式.方法的思想是通过最小化图像邻域的概率密度函数的熵来给出最优分割.熵的最小化使用了一种快速的水平集方案.这种方法并不依赖于学习阶段的数据,是无监督的.根据数据的信息内容自动调整内部一些重要参数.

This paper presents a novel approach to unsupervised texture segmentation according to a very general nonparametric statistical model of image neighborhoods. The method models image neighborhoods directly without the constuction of intermediate features. It is a generic apporach that tries to adapt to a variety of textures. It exploits the fundamental description of testures as images dedved from stationary random fidds and models the associated higher- order statistics nonparametrically. The method minimizes an entropy-based metric on the probability density functions of image neighborhoods to give an optimal segmentation. The entropy minimization drives a very fast level-set scheme that uses threshold dynamics, which allows for a very rapid evolution towards the optimal segmentation during the initial iterations. The method does not rely on a training stage and, hencc, is unsupervised. It adjusts its important internal parameters automatically based on the content of the data.