MRF model and FRAME model-based unsupervised image segmentation

被引：0

作者：

CHENG Bing WANG Ying ZHENG Nanning JIA Xinchun BIAN Zhengzhong Institute of Artificial Intelligence and Robotics Xian Jiaotong University Xian China ^{[710049
]}

Department of Biomedical Engineering Xian Jiaotong University Xian China ^{[710049
]}

The Research Center of The First Hospital Xian Jiaotong University Xian China Correspondence should be addressed to Cheng Bing ^{[710061
]}

机构：

来源：

ScienceinChina(SeriesF:InformationSciences) | 2004年 / 06期

关键词：

image segmentation; Markov random field; FRAME model; Maximum a Posterior estimation; iterated conditional modes;

D O I：

暂无

中图分类号：

TP391.41 [];

学科分类号：

080203 ;

摘要：

<正>This paper presents a method for unsupervised segmentation of images consisting of multiple textures. The images under study are modeled by a proposed hierarchical random field model, which has two layers. The first layer is modeled as a Markov Random Field (MRF) representing an unobservable region image and the second layer uses "Filters, Random and Maximum Entropy (Abb. FRAME)" model to represent multiple textures which cover each region. Compared with the traditional Hierarchical Markov Random Field (HMRF), the FRAME can use a bigger neighborhood system and model more complex patterns. The segmentation problem is formulated as Maximum a Posteriori (MAP) estimation according to the Bayesian rule. The iterated conditional modes (ICM) algorithm is carried out to find the solution of the MAP estimation. An algorithm based on the local entropy rate is proposed to simplify the estimation of the parameters of MRF. The parameters of FRAME are estimated by the ExpectationMaximum (EM) algorithm. Finally, an exp

引用

页码：697 / 705

页数：9

共 3 条

[1] Variation-based approach to image segmentation [J].

Yongping Zhang ;

Nanning Zheng ;

Rongchun Zhao .

Science in China Series : Information Sciences, 2001, 44 (4) :259-269

[2]

Maximum-entropy clustering algorithm and its global convergence analysis[J] . Zhihua Zhang,Nanning Zheng,Gang Shi.Science in China Series E: Technological Sciences . 2001 (1)

[3]

Filters, Random Fields and Maximum Entropy (FRAME): Towards a Unified Theory for Texture Modeling[J] . Song Chun Zhu,Yingnian Wu,David Mumford.International Journal of Computer Vision . 1998 (2)

← 1 →