In this paper, we propose a finite mixture of inhomogeneous Markov random fields for image segmentation. Prior probabilities of pixel class memberships are condi- tionally specified through a Gibbs distribution, where the association between pixel labels is modeled by a class-specific term. We show how model parameters can be estimated in a pseudo maximum likelihood framework.
Finite mixture models for image segmentation
Nieddu L;
2005-01-01
Abstract
In this paper, we propose a finite mixture of inhomogeneous Markov random fields for image segmentation. Prior probabilities of pixel class memberships are condi- tionally specified through a Gibbs distribution, where the association between pixel labels is modeled by a class-specific term. We show how model parameters can be estimated in a pseudo maximum likelihood framework.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
