Image Restoration, Modelling, and Reduction of Dimensionality

Abstract

Recursive restoration of two-dimensional noisy images gives dimensionality problems leading to large storage and computation time requirements on a digital computer. This paper shows a two-dimensional second-order Markov process representation can be used for fast recursive restoration of images with small storage requirements. Advantages of this method over existing techniques are illustrated by means of examples.