RD-Optimization of Hierarchical Structured Adaptive Vector Quantization for Video Coding
DCC '00 Proceedings of the Conference on Data Compression
| Hi-index | 0.00 |
We present a video encoding scheme based on mean-removed vector quantization where the shape codebook is adapted to the source on a frame-by-frame basis. In contrast to previously published adaptive vector quantization (AVQ) approaches the decision how blocks are encoded and how the codebook is updated is done in a rate-distortion optimized fashion. Furthermore, we show that the translation of our rate-distortion optimized AVQ into the wavelet domain leads to an improved coding performance. We present coding results as well as an empirical analysis of the short- and long-term behavior of the adaptive codebook.