Probabilistic image processing by extended Gauss-Markov random fields

Kazuyuki Tanaka, Nicolas Morin, Muneki Yasuda, D. M. Titterington

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We propose an extension of the Gauss-Markov random field (GMRF) models by introducing next-nearest neighbour interactions. The values of the next-nearest neighbour interactions are set to positive real numbers with the expectation that this will lead to some noise reduction while preserving the edges. Values for the hyperparameters in the proposed model are determined by using the EM algorithm in order to maximize the marginal likelihood. In addition, a measure of mean squared error, which quantifies the statistical performance of our proposed model, is derived analytically as an exact explicit expression by means of the multi-dimensional Gaussian integral formulas.

Original languageEnglish
Title of host publication2009 IEEE/SP 15th Workshop on Statistical Signal Processing, SSP '09
Pages618-621
Number of pages4
DOIs
Publication statusPublished - 2009 Dec 25
Event2009 IEEE/SP 15th Workshop on Statistical Signal Processing, SSP '09 - Cardiff, United Kingdom
Duration: 2009 Aug 312009 Sep 3

Publication series

NameIEEE Workshop on Statistical Signal Processing Proceedings

Other

Other2009 IEEE/SP 15th Workshop on Statistical Signal Processing, SSP '09
CountryUnited Kingdom
CityCardiff
Period09/8/3109/9/3

Keywords

  • Bayesian image analysis
  • Bayesian network
  • Gauss-Markov random fields

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Applied Mathematics
  • Signal Processing
  • Computer Science Applications

Fingerprint Dive into the research topics of 'Probabilistic image processing by extended Gauss-Markov random fields'. Together they form a unique fingerprint.

Cite this