Motion optimized spatial-temporal video coding based on wavelet transform

For the very low bit-rate video compression, wavelet based approaches suffer the ghost effect caused by larger temporal distortions. To solve this problem, we propose a motion optimized spatial-temporal coding algorithm, which has two independent components: 1) the sub-grouping transformation (SGT) algorithm which divides a coefficient array into temporal sub-groups and applies the temporal direction wavelet transform in sub-groups respectively; 2) the intra-band discriminating quantization (IBDQ) algorithm which divides a group of frames into spatial motion areas and assigns different quantization steps for these motion areas according to their motion intensities. By spatially assigning large quantization steps for slow motion areas of the video and temporally splitting high motion coefficient arrays into smoother sub-groups, the proposed algorithms can achieve high compression ratios and reduce the ghost effect at the same time.