Realizations and parameter estimation for line processes

The Markov line process that has been used in some image segmentation and restoration studies is investigated. Realizations from this model are presented for a wide range of parameter values, and the effects of certain parameters are studied. The maximum pseudolikelihood (MPL) estimation procedure is implemented for the Markov line process. The MPL procedure is applied to several images generated from the model as well as to a hand-drawn image and the edge-detector output of a natural image. It is expected that improved segmentation and restoration results can be obtained, if the Markov line process model is fine-tuned to the class of images under consideration, by estimating the parameters of some typical images in that class.<<ETX>>

[1]  Chee Sun Won,et al.  Maximum likelihood estimation of Gaussian Markov random field parameters , 1988, ICASSP-88., International Conference on Acoustics, Speech, and Signal Processing.

[2]  Donald Geman,et al.  Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  Haluk Derin,et al.  Modeling and Segmentation of Noisy and Textured Images Using Gibbs Random Fields , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  John W. Woods,et al.  Compound Gauss-Markov random fields for image estimation , 1991, IEEE Trans. Signal Process..