Three dimensional wavelet based approach for a scalable video conference system

In this paper a video coder based on a spatiotemporal multiresolution pyramid generated by a 3D separable wavelet transform (WT) is presented. The aim of the proposed coding scheme is to provide an effective scalability of the user profile by organizing the full resolution encoded bit stream into substreams, one for each temporal layer of the pyramid, which are then partitioned into substreams at increasing spatial resolution. This allows to simultaneously distribute, at different time and space resolutions, the same video to multiple users accessing the videoconference services through channels of different capacity, without multiple encoding.