IEEE TRANSACTIONS ON CONSUMER ELECTRONICS, cilt.53, sa.3, ss.1125-1129, 2007 (SCI-Expanded)
Post-filtering is an effective tool for improvement of encoded image and video quality at low bit-rates. Some consumer electronics devices such as mobile phones apply excessive compression to meet limited bandwidth requirements. Thus, some annoying visual artifacts occur in the decoded image/video. This paper presents a deblocking method based on fuzzy logic to alleviate one of the most common visual artifacts (i.e. blocking) encountered in heavily compressed images. The proposed approach takes intensity difference and variance of pixels at the block boundaries into account in a Juz v manner to decide amount of blocking effect. Once the proposed approach decides the type of the blockiness, a simple interpolation according to the strength is carried out. Experimental results show the effectiveness of the proposed approach.