A Method Using Nonparametric Hidden Markov Trees for Image Denoising

Author

Song, Wang ; Weihong, Wang

Author_Institution

Coll. of Software, Zhejiang Univ. of Technol., Hangzhou, China

Volume

1

fYear

2009

fDate

6-7 June 2009

Firstpage

143

Lastpage

146

Abstract

A hierarchical, nonparametric statistical model for wavelet representations of natural images is developed in this paper. Extending previous work on Gaussian scale mixtures, wavelet coefficients are marginally distributed according to infinite, Dirichlet process mixtures. A hidden Markov tree is then used to couple the mixture assignments at neighboring nodes. Via a Monte Carlo learning algorithm, the resulting hierarchical Dirichlet process hidden Markov tree (HDP-HMT) model automatically adapts to the complexity of different images and wavelet bases. Image denoising results demonstrate the effectiveness of this learning process.

Keywords

Gaussian processes; Monte Carlo methods; hidden Markov models; image denoising; trees (mathematics); Dirichlet process mixtures; Gaussian scale mixtures; Monte Carlo learning algorithm; hierarchical Dirichlet process-hidden Markov tree model; hierarchical statistical model; image denoising; learning process; nonparametric hidden Markov trees; nonparametric statistical model; wavelet representations; Bayesian methods; Computational intelligence; Educational institutions; Frequency; Gaussian distribution; Hidden Markov models; Image denoising; Monte Carlo methods; Statistical distributions; Wavelet coefficients; hidden Markov trees; image denoising; nonparametric Bayesian methods;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Intelligence and Natural Computing, 2009. CINC '09. International Conference on

Conference_Location

Wuhan

Print_ISBN

978-0-7695-3645-3

Type

conf

DOI

10.1109/CINC.2009.49

Filename

5231221