| 摘要 |
Provided is a methodextracts a merge index from a bit stream constructs a merge candidate list using available spatial and temporal merge candidates selects a merge predictor among merge candidates listed in the merge candidate list using the merge index and sets motion information of the merge predictor as motion information of the current prediction unit. The temporal merge candidate includes a reference picture index and a motion vector zero is set as the reference picture index of the temporal merge candidate and a motion vector of a temporal merge candidate block of a temporal merge candidate picture is set as the motion vector of the temporal merge candidate. Accordingly the coding efficiency of the motion information is improved by including various merge candidates. Also the computational complexity of an encoder and a decoder is reduced maintaining improvement of coding efficiency by adaptively storing motion information of reference picture and adaptively generating a temporal merge candidate. |
