Optimal nonlinear estimation on the plane (Q1082324)

Regarding the rectangular observation domain as a realization of a vector random process, a formulation of optimal nonlinear estimates in a class of partially observable Markov random fields is proposed. This approach makes it possible to surmount a number of difficulties typically encountered in conventional estimation procedures. The method discussed here arises in a natural way in the design of real-time environmental monitoring systems as remote detection and demarcation of forest-fire zones, oil spilles on the ocean surface, etc.