BAYESIAN IMAGE PROCESSING OF DATA FROM CONSTRAINED SOURCE DISTRIBUTIONS.

Title: BAYESIAN IMAGE PROCESSING OF DATA FROM CONSTRAINED SOURCE DISTRIBUTIONS.
Identifier: AAI8708302
identifier: 8708302
Creator: LIANG, ZHENGRONG.
Contributor: Hiram Hart
Date: 1987
Language: English
Publisher: City University of New York.
Subject: Biophysics, Medical
Abstract: A new Bayesian image processing (BIP) formalism which incorporates various categories of a priori source distribution information in treating measured data which obeys poisson or gaussian statistics is introduced and two classes of multiplicative and additive forms of iterative BIP algorithms are formulated. Different categories of a priori source information are described and the resulting probability source distributions developed.;Practical application of this work falls in the areas of biomedical and space related image reconstruction and restoration. Specifically BIP is most effective in those situations in which the a priori source information can be characterized in terms of probability functions of the source distributions and the noise in the measured data would otherwise significantly obscure the resulting images.;The results obtained using the BIP formalism on computer generated and experimental data from phantoms are clearly superior to standard methods wherever the a priori source information is reasonably accurate and sufficiently restrictive.
Type: dissertation
Source: PQT Legacy CUNY.xlsx
degree: Ph.D.
Program: Physics

Media: BAYESIAN IMAGE PROCESSING OF DATA FROM CONSTRAINED SOURCE DISTRIBUTIONS.