Scalable coding of subband images with quadtree-based classified vector quantization

Scalable coding is useful when a variety of resolutions are simultaneously required or when resilience to errors is necessary. This paper describes spatial and SNR scalable coding schemes which are based on subband coding with quadtree-based classified vector quantization. It is expected that the proposed scalable coding techniques can be efficiently adapted for simulcast coding and error resilient coding over ATM networks when used with interframe coding and motion compensation techniques. It is also shown that the proposed techniques are superior to the JPEG standard in which hierarchical (spatial scalability) and progressive (SNR scalability) modes are used for the comparison.