Video encoding method and apparatus, and video decoding apparatus
To allow a finer quantization control according to the property of an image within a macroblock, quantization parameter values are allowed to be changed in units of sub-blocks equal to or smaller than the macroblock in a similar manner as in motion compensation and orthogonal transform processes. A finer-tuned quantization control is performed, for example, by selecting fine and coarse quantization parameters respectively for corresponding sub-blocks if a plurality of images having different properties coexist within the macroblock.
1. A video encoding method for dividing each frame of a video into blocks of a predetermined size, and for encoding each of the blocks by combining temporal compression using pixel difference data and motion compensation, an orthogonal transform for transforming spatial information into frequency information, and quantization of an orthogonal transform coefficient, by using a quantization parameter defined corresponding to the blocks of the predetermined size, the video encoding method comprising:
further dividing the blocks of the predetermined size into sub-blocks of a size smaller than the predetermined size;
deciding a quantization parameter for each of the sub-blocks according to an image property;
quantizing the orthogonal transform coefficient with the quantization parameter for each of the sub-blocks; and
encoding the decided quantization parameter, wherein
each of the sub-blocks for which the quantization parameter is decided has a size equal to or larger than a block size that is a unit of processing in the orthogonal transform, and
an intermediate value of quantization parameters of blocks of the predetermined size or sub-blocks is calculated for the blocks of the predetermined size or the sub-blocks that are encoded and neighboring an encoding sub-block, and the quantization parameter for each of the sub-blocks is encoded by encoding a difference between the intermediate value of quantization parameters of a plurality of encoded blocks adjacent to the encoding sub-block and a quantization parameter of the encoding sub-block.
2. The video encoding method according to claim 1 , wherein
the orthogonal transform is performed in units of the sub-blocks into which the blocks of the predetermined size are further divided, and
a division shape of the sub-blocks for the quantization parameter is made identical to a division shape of the sub-blocks in the orthogonal transform.
3. The video encoding method according to claim 1 , wherein
the motion compensation is performed in units of the sub-blocks into which the blocks of the predetermined size are further divided, and
a division shape of the sub-blocks for the quantization parameter is made identical to a division shape of the sub-blocks for the motion compensation.