Convergence of unsupervised image segmentation algorithms

Research output: Contribution to journalConference articlepeer-review

2 Scopus citations

Abstract

This paper presents a comparative study of three deterministic unsupervised image segmentation algorithms. All of the three algorithms basically make use of a Markov random field (MRF) and try to obtain an approximate solution to the maximum likelihood (ML) or the maximum a posteriori (MAP) estimates. Although the three algorithms are based on the same stochastic image models, they adopt different ways to incorporate model parameter estimation into the iterative region label updating procedure. The differences among the three algorithms are identified and the convergence properties are compared both analytically and experimentally.

Original languageEnglish
Pages (from-to)209-220
Number of pages12
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume2568
DOIs
StatePublished - 11 Aug 1995
EventNeural, Morphological, and Stochastic Methods in Image and Signal Processing 1995 - San Diego, United States
Duration: 9 Jul 199514 Jul 1995

Keywords

  • Expectation and maximization
  • Gibbs random fields
  • Iterated conditional modes
  • Markov random fields
  • Maximum a pesteriori estimation
  • Maximum likelihood estimation
  • Unsupervised image segmentation

Fingerprint

Dive into the research topics of 'Convergence of unsupervised image segmentation algorithms'. Together they form a unique fingerprint.

Cite this