A novel blocking map guided adaptive fuzzy deblocking filter

Abstract

This paper proposes a novel blocking map guided fuzzy filter to remove blocking artifacts in compressed images. The fuzzy filter is controlled adaptively by a blocking artifact map. The blocking artifact map is built on local characteristics of the image content. The masking effect of human visual system (HVS) is also considered to estimate the blocking artifact level more accurately. Experimental results show that blocking artifact map fits well to the visual system. The proposed deblocking method has better detail preservation and lower computational costs than other methods. It achieves better performance in both reducing noise and maintaining better texture detail between block borders than the other algorithms.