Parametric image segmentation consists of finding a label field that defines a partition of an image into a set of nonoverlapping regions and the parameters of the models that describe the variation of some property within each region. A new Bayesian formulation for the solution of this problem is presented, based on the key idea of using a doubly stochastic prior model for the label field, which allows one to find exact optimal estimators for both this field and the model parameters by the minimization of a differentiable function. An efficient minimization algorithm and comparisons with existing methods on synthetic images are presented, as well as examples of realistic applications to the segmentation of Magnetic Resonance volumes and to motion segmentation.

No abstract

No abstract

No reproduction, copy or transmission may be made without written permission from the individual authors. Review ProcessPapers submitted to this conference have been double-blind peer reviewed before final acceptance to the conference. Initially, abstracts were reviewed for relevance and accessibility and successful authors were invited to submit full papers. Many thanks to the reviewers who helped ensure the quality of all the submissions. Ethics and Publication Malpractice PolicyACIL adheres to a strict ethics and publication malpractice policy for all publicationsdetails of which can be found here: http://www.academic-conferences.org/policies/ethics-policy-for-publishing-in-theconference-proceedings-of-academic-conferences-and-publishing-international-limited/ Conference ProceedingsThe Conference Proceedings is a book published with an ISBN and ISSN. The proceedings have been submitted to a number of accreditation, citation and indexing bodies including Thomson ISI Web of Science and Elsevier Scopus.Author affiliation details in these proceedings have been reproduced as supplied by the authors themselves.