Method for generating prediction block in AMVP mode
A method for generating a prediction block in Advanced Motion Vector Prediction (AMVP) mode to reconstruct a prediction-coded video signal using a motion vector approximate to original motion information. An AMVP candidate list is generated using effective spatial and temporal AMVP candidates for a current Prediction Unit (PU). If the number of the effective AMVP candidates is smaller than a predetermined value, a motion vector having a predetermined value as a candidate to the AMVP candidate list. Then a motion vector corresponding to an AMVP index of the current PU from among motion vectors included in the AMVP candidate list is determined to be a motion vector predictor of the current PU.
1. A method of encoding an image in a merge mode, the method comprising:
determining motion information of a current prediction unit, and generating a prediction block using the motion information;
generating a residual block using an original block and the prediction block, transforming the residual block to generate a transformed block, quantizing the transformed block using a quantization parameter to generate a quantized block, and scanning the quantized block to entropy-code the quantized block; and
encoding the motion information using effective spatial and temporal merge candidates of the current prediction unit,
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 the quantization parameter is encoded using an average of two effective quantization parameters among a left quantization parameter, an upper quantization parameter and a previous quantization parameter of a current coding unit, and
wherein when the quantized block is larger than a predetermined size, the quantized block is divided into a plurality of subblocks to be scanned, and a scan pattern for scanning the plurality of subblocks is the same as a scan pattern for scanning quantized coefficients within each subblock.
2. The method of claim 1 , wherein the plurality of subblocks and the quantized coefficients are scanned in a reverse direction.
3. The method of claim 1 , wherein the temporal merge candidate picture is determined according to a slice type.
4. The method of claim 3 , wherein the temporal merge candidate block is a first effective block when a first candidate block and a second candidate block are retrieved in this order.
5. The method of claim 4 , wherein the first candidate block is a bottom-left block adjacent to a block corresponding to the current prediction unit within the temporal merge candidate picture, and the second candidate block is a block including a upper-left pixel of a center of the block corresponding to the current prediction unit within the temporal merge candidate picture.
6. The method of claim 1 , wherein, if only one of the left quantization parameter, the upper quantization parameter and the previous quantization parameter is effective, the quantization parameter of the current coding unit is encoded using the effective quantization parameter.