Mean-field theory for grayscale texture synthesis using Gibbs random fields

I. M. Elfadel, A. L. Yuille

Research output: Contribution to journalConference articlepeer-review

4 Scopus citations

Abstract

In this paper, we show how methods developed in the context of statistical physics can be used to analyze grayscale texture synthesis procedures based on the probabilistic paradigm of Gibbs random fields (GRF). In particular, we show, using the mean-field equations of the texture GRF, the existence of a bifurcation point indicating the presence of a phase transition in the textural pattern. Using simulations, we show that a number of interesting phenomena occur at the bifurcation temperature like a sudden decrease in energy, sharp peaks in similarity measures between textural patterns, and sudden saturation of the mean-field variables. We argue that for texture synthesis, it is sufficient to simulate the mean-field equations near the bifurcation temperature.

Original languageBritish English
Pages (from-to)248-259
Number of pages12
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume1569
DOIs
StatePublished - 1 Oct 1991
EventStochastic and Neural Methods in Signal Processing, Image Processing, and Computer Vision 1991 - San Diego, United States
Duration: 21 Jul 1991 → …

Fingerprint

Dive into the research topics of 'Mean-field theory for grayscale texture synthesis using Gibbs random fields'. Together they form a unique fingerprint.

Cite this