| 摘要 |
PROBLEM TO BE SOLVED: To provide a favorable video decoding method.SOLUTION: Disclosed is a method of decoding video data in a merge mode including: generating a merge candidate list by utilizing available spatial merge candidates and temporal merge candidates; deriving motion information by utilizing a merge index and the merge candidate list; generating a prediction block by utilizing the motion information; and generating a residual block by inverse-quantizing the quantized block by utilizing a quantization parameter and a quantization matrix and by inverse-transforming the inverse quantized block. The quantization parameter is generated per quantization unit, and a minimum size of the quantization unit is adjusted per picture. The coding efficiency of motion information is improved by selecting one of neighboring blocks and a temporal candidate block as a motion information predictor. The coding efficiency of residual block is improved by adaptively adjusting the quantization parameter per quantization unit and generating a quantization parameter predictor by utilizing multiple neighboring quantization parameters. |
