Learning structural and corruption information from samples for Markov random field binary image reconstruction

The authors have advanced Markov random field research by addressing the issue of obtaining a reasonable, nontrivial, noise model. They address this issue by looking at original images together with noisy imagery, and so creating a probability distribution for pairs of neighborhoods across both images. This models the noise within the MRF probability distribution, and provides an easy way to generate Markov random fields for annealing or other relaxation methods.<
>