Mathematical model of nongaussian vectorial signals and images

In paper the approach to modeling heterogeneous signals and images is described. The example of multilevel piece-constant and binary images description is presented. For the description of the heterogeneous, phenomena, the <> random binary Markov process is used. Hence, procedure of detection of heterogeneous features is reduced to the filtration of <> signal. Results of the real images processing, based on the submitted models is presented.