Issue No. 09 - September (1989 vol. 11)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/34.35503
<p>The authors present the class of semi-Markov random fields and deal, in particular, with the subclass of discrete-valued, nonsymmetric half-plane, unilateral Markov random fields. A hierarchical nonstationary-mean nonstationary-variance (NMNV) image model is proposed for the modeling of nonstationary and noisy images. This model seems to be advantageous as compared to a regular NMNV model because it statistically incorporates the correlation between pixels around the boundary of two adjacent regions. The hierarchical NMNV model leads to the development of an optimal algorithm that allows the simultaneous segmentation and model estimation of measured images. Although no theoretical result is available for the consistency of the estimated model parameters, the method seems to work sufficiently well for the examples considered.</p>
Markov random fields; picture processing; model estimation; nonstationary noisy images; hierarchical nonstationary-mean nonstationary-variance; image model; correlation; simultaneous segmentation; Markov processes; optimisation; parameter estimation; picture processing
J. Mendel and J. Goutsias, "Simultaneous Optimal Segmentation and Model Estimation of Nonstationary Noisy Images," in IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 11, no. , pp. 990-998, 1989.