Extended time motion vector prediction for scalable video coding

摘要

enhancement layer: In the (EL enhancement layer) coding, time candidates can be derived in the merge mode and non-merge mode. The current prediction unit (PU: prediction unit) of the reference choice of the index, and, collocated is (co-located) selection of PU of the reference list, jointly be considered, of EL for the merge mode time It can be carried out when building the candidates. Selection may be performed as the motion vector (MV) scaling operations may be avoided. Motion vector prediction candidates to be selected, the merging candidate list may be added at the position of the previous one spatial motion vector prediction candidates. Motion vector prediction candidates to be selected, the merging candidate list may be added in place of the spatial motion vector prediction candidates.