Successive structural analysis using wavelet transform for blocking artifacts suppression (Q5958156)

A successive structural analysis using wavelet transforms is proposed to remove the blocking artifacts associated with block-based coding techniques such as JPEG. The successive structural analysis is based on a study of the magnitude of the wavelet coefficients using a set of threshold values. This allows the structural information in an image and the information about the block boundaries to be separated into several classes. A metric is then used to quantify the quality of the reconstruction from each class. During reconstruction, only those classes that correspond to the structural information are retained while classes that correspond to the block boundaries are discarded. Simulation results show that our proposed algorithm can effectively suppress the blocking artifacts without over-smoothing the image details.