Embedded multiple subbands scaling image coding using reversible integer wavelet transforms
Libao Zhang, Ke Wang · 2005
The integer wavelet transform (IWT) and IWT-based image coding offer many advantages over the traditional wavelet. But the coefficient magnitude of the IWT-image have smaller dynamic change value and worse energy compaction than the first generation wavelet. These reduce the efficiency of IWT-based coding. In this paper, a new coding algorithm so-called embedded multiple subbands scaling coding (EMSSC) is presented based on reversible IWT. The presented method exploits two coding strategies - multiple subbands scaling (MSS) and fast quadtree partitioning (FQP). During the transform, the scaling parameter of each band is chosen for optimizing the transform coefficient distribution of each subband. During coding, all image coefficients are encoding using a simple, efficient quadtree partitioning method. Simulation results show that the EMSCC algorithm provides PSNR performance up to 0.6/spl sim/1.2 dB better than SPIHT and SPECK algorithm using IWT and is comparable to SPIHT and SPECK using DWT of Daubechies 9/7 wavelet filters.