On the optimality of a new class of 2D recursive filters

The purpose of this paper is to prove the minimum variance property of a ne w class of 2D, recursive, finite-dimensional filters. The filtering algorit hms are derived from general basic assumptions underlying the stochastic mo delling of an image as a 2D gaussian random field. An appealing feature of the proposed algorithms is that the image pixels are estimated one at a tim e; this makes it possible to save computation time and memory requirement w ith respect to the filtering procedures based on strip processing. Experime ntal results show the effectiveness of the new filtering schemes.