The control of quantization step size after DCT(discrete cosine transform) is the key process for data compression in the motion-compensated video coding. In general, it is a weighted quantization for intraframe signals and unweighted quantization for interframe difference signals.
In this thesis, a common structure of weighted quantization for both type of input signals is proposed. It is shown that by modifying the DPCM loop, the same weighting matrix for intraframe image can be used also for interframe difference image. It is found that the subjective quality of the proposed scheme is better than the conventional one, particularly for the regions of less details.
