Title :
A new approach to parameter estimation for Gibbs random fields
Author :
Derin, H. ; Elliott, H. ; Kuang, J.
Author_Institution :
University of Massachusetts, Amherst, Massachusetts
Abstract :
This paper presents a new approach for estimating the parameters of random fields characterized by Gibbs distributions. Unlike the commonly used "coding method", the parameter estimation procedure presented here involves the solution of a set of linear equations. We also present numerical and pictorial results on application of this estimation procedure on several Gibbs distributed images.
Keywords :
Application software; Computer vision; Image coding; Image processing; Lattices; Markov random fields; Maximum likelihood estimation; Nonlinear equations; Parameter estimation; Yield estimation;
Conference_Titel :
Acoustics, Speech, and Signal Processing, IEEE International Conference on ICASSP '85.
DOI :
10.1109/ICASSP.1985.1168296
