Ringing artifact reduction of JPEG images using a SGLI prior

This paper deals with the ringing artifact problem of JPEG compressed images. The ringing artifacts occur because of loss of high frequency components and appear around sharp edges in images. They degrade the quality of picture and cause unpleasant viewing experiences to viewers. Therefore, we propose a novel Bayesian method which removes the ringing artifacts efficiently in JPEG images. The proposed method employs a Bayesian framework based on a SGLI (spatial gradient and local inhomogeneity) prior. The SGLI prior is from two discontinuity measures which have been utilized as an illumination normalization method for robust face recognition. Experimental results show that the proposed method gives an averaging 0.21 dB gain in PSNR for the JPEG compression artifact reduction.