PUMA
Istituto di Scienza e Tecnologie dell'Informazione     
Tonazzini A., Bedini L. Degradation identification and model parameter estimation in discontinuity-adaptive visual reconstruction. In: Advances in Imaging and Electron Physics, vol. 120 pp. 193 - 284. Elsevier, 2002.
 
 
Abstract
(English)
This chapter describes the progress made in finding efficient solutions to the highly ill-posed and computationally demanding problem of blind and unsupervised visual reconstruction. The problem was dealt with in the general framework of edge-preserving regularization through the use of Bayesian estimation and Markov random field (MRF) image models. This approach is known to be one of the most promising and efficient for solving a large body of problems in image processing and computer vision. The chapter also describes a fully Bayesian approach that is essentially based on the joint maximization of a distribution of the image field, the data, and the degradation and model parameters. This very complex joint maximization was initially decomposed into a sequence of maximum a posteriori (MAP) and/or maximum likelihood (ML) estimations, to be performed alternately and iteratively, with an initial significant reduction in complexity and computational load. The saddle point approximation from the statistical mechanics and the importance sampling theorem from the Markov chain Monte Carlo (MCMC) theory were then applied to further improve the computational performance of the MRF parameter estimation.
URL: http://biblioproxy.cnr.it:2052/science/article/pii/S1076567002800362
DOI: 10.1016/S1076-5670(02)80036-2
Subject



Icona documento Open access Icona documento Restricted Icona documento Private

 


Per ulteriori informazioni, contattare: Librarian http://puma.isti.cnr.it

Valid HTML 4.0 Transitional