Method of constructing merge list
An image decoding method using a merge mode, can include constructing a merge candidate list by using available spatial merge candidates and an available temporal merge candidate; generating a prediction block based on motion information derived by using the merge candidate list; inversely quantizing a quantization block by using a quantization parameter generated per quantization unit, wherein the quantization block is generated with quantization coefficient components; generating a residual block by inversely transforming the inversely quantized quantization block; deriving a reconstructed block by using the prediction block and the residual block, in which the spatial merge candidates are motion information of spatial merge candidate blocks, and the spatial merge candidate blocks are a left block, an above block, an above-right block, a left-below block or an above-left block of a current block, and a reference picture index of the temporal merge candidate is 0.
1. An image decoding method using a merge mode, the method comprising:
constructing a merge candidate list by using available spatial merge candidates and an available temporal merge candidate;
generating a prediction block based on motion information derived by using the merge candidate list;
inversely quantizing a quantization block by using a quantization parameter generated per quantization unit, wherein the quantization block is generated with quantization coefficient components;
generating a residual block by inversely transforming the inversely quantized quantization block;
deriving a reconstructed block by using the prediction block and the residual block,
wherein the spatial merge candidates are motion information of spatial merge candidate blocks, and the spatial merge candidate blocks are a left block, an above block, an above-right block, a left-below block or an above-left block of a current block,
wherein a reference picture index of the temporal merge candidate is 0,
wherein the quantization parameter is generated using a differential quantization parameter and a quantization parameter predictor, and the quantization parameter predictor is generated by using a left quantization parameter and an above quantization parameter if the left and the above quantization parameters are available,
wherein, when only one of the left quantization parameter and the above quantization parameter is available, the quantization parameter is generated using the available quantization parameter and a previous quantization parameter, and
wherein, when the left and above quantization parameters are not present, the previous quantization parameter is set as the quantization parameter predictor.
2. The method of claim 1 , wherein the quantization unit includes at least one coding unit, and a size of the quantization unit is adjusted per a picture.
3. The method of claim 1 , wherein each of the quantization unit includes the differential quantization parameter.
4. The method of claim 1 , wherein the quantization parameter predictor is generated per the quantization unit.
5. The method of claim 1 , wherein one or more candidates are added in the merge candidate list if a number of available spatial and temporal merge candidates is smaller than a predetermined number.
6. The method of claim 1 , if a current coding unit has an allowed minimum size, asymmetric partitioning is not allowed.
7. The method of claim 1 , wherein a motion vector of the temporal merge candidate is a motion vector of a temporal merge candidate block within a temporal merge candidate picture, and a position of the temporal merge candidate block is determined depending on a position of a current prediction unit within an LCU.