Hybrid coding of video with spatio-temporal scalability using subband decomposition

The paper deals with scalable (hierarchical) coding of video at bitrates in the range of about 2-5 Mbps. A practical goal is to obtain comparable bitrates in a base layer and an enhancement layer. The solution proposed in the paper is based on both temporal and spatial resolution reduction performed for data transmitted in a base layer. Two principal variants are presented: systems with three-dimensional filter banks for spatio-temporal analysis and systems where some B-frames are allocated to an enhancement layer but subband analysis is purely spatial. The assumption is that a base layer is fully MPEG-2 compatible. Bitstreams related to high spatial frequencies are encoded either using DCT-based MPEG-like coding or a technique that exploits mutual dependencies between low- and high-spatial-frequency subchannels.