Combined frequency and spatial domain algorithm for the removal of blocking artifacts

Abstract

A novel combined frequency and spatial domain method is presented in this paper for blockiness reduction for low bit rate compressed images. The method consists of two stages: in the first, better estimates of the reconstructed DCT coefficients are obtained based on their observed probability distribution. In the second, an efficient postprocessing scheme consisting of a region classification algorithm and a spatial adaptive filtering is applied for blockiness removal. The type of filtering is decided on the basis of an estimation of the local characteristics of the coded image. The main advantage of the proposed method is the efficient combination and design of these two stages which are acting complementarily for the reduction of blocking artifacts. This approach is shown to produce excellent results in removing blocking artifacts. The efficient performance of the proposed algorithm is due, firstly, to the proposition that the shape and the position of the filter kernel are adjusted according to the characteristics of the local image region and secondly, to the employment of the modified improved DCT coefficients by the postprocessing filter. Experimental results illustrating the performance of the proposed method are presented and evaluated.