Stochastic Systems

On solution of the two-dimentional digital filtering problem (design of matrix filters)

V. M. Kuntsevich

Kiev

Abstract: Two-dimensional digital fields are to be filtered and extrapolated whose states at discrete times are described by appropriate state matrices. A matrix analog of the discrete-time random Markov process is introduced and it is shown that if the sequence of matrices is a process of this type, then the optimal filtering reduces to obtaining a matrix dialog of the Kalman filter. The extrapolation is also obtained by using a matrix analog of autoregression control.

UDC: 519.711.2


Received: 03.02.1986