| 摘要 |
In a high efficiency encoding system, a picture is divided into many three-dimensional blocks, or blocks each consisting of n areas in each of n frames continuous in terms of time. Any movement of the picture is detected in each block to develop movement information. The picture element number of each block is sub-sampled, and the sub-sampling phases are sequentially shifted with respect to the blocks. Data of picture elements thinned by sub-sampling (i.e., not transmitted) are decoded, while data relating to a still portion of the picture is decoded with excellent results by merely combining sub-sampling patterns.
|
