Fuzzy edge detector based blocking artifacts removal of DCT compressed images

DCT compressed digital images vulnerable to block like visible distortions at low bit rate. To overcome these block distortions, a new compressed image artifact removal method as cascade of fuzzy edge detector and fuzzy based substitution is proposed. In this scheme, Gaussian type fuzzy edge detector is used for each overlapped block of a DCT compressed image. The central pixel of compressed image block is to be considered under block edge boundary depending on fuzzy edge detector and its value is substituted with combination of mean, median values of its neighbors and S-shaped fuzzy membership function values. Presented experimental results illustrate the performance of proposed algorithm in terms of visual appearance and also in terms of quantified values as MSE, PSNR.