Fully scalable video codec

In this paper, we present the video coder which achieves various kinds of continuous scalabilities simultaneously, including spatial scalability, temporal scalability and SNR scalability. By applying the wavelet transform that maps integer to integer in spatial and temporal dimensions, it is possible to perfectly reconstruct the original video sequences. Due to the energy compaction in the temporal domain, using wavelet transform alone does not work well, we replace it by applying the motion estimation at the Low-Low (LL) band to eliminate ringing effects. Following the 2-D SPECK algorithm to represent the wavelet (or residual) coefficients based on the bit plane quantization (according to magnitude comparisons), a precise rate control and a self- adjusting rate allocation are achieved. In order to provide all kinds of scalabilities, a multiresolution encoding/decoding concept is presented and implemented. In which the significant wavelet coefficients corresponding to different resolutions in the original embedded bit stream are interleaved. In the multiresolution decoding process, the user can specify his (or her) own desired spatial and temporal resolutions, and the server will produce the combine bit stream according to the user requirement. In addition to spatial and temporal scalabilities, SNR scalability can be achieved by fitting the exact target bitrate to the available network bandwidth.