Inter-Picture Prediction in HEVC - High Efficiency Video Coding (HEVC) - page 116

CTU0 CTU1 MB0 MB1

Fig. 5.3 Maximum number of left neighbors with motion data in HEVC with CTU0 having 16

8 4 luma PBs next to CTU1 with one 64 64 luma PB ( left ) and in H.264/AVC with MB0 having

four 4 4 partitions next to MB1 with one 16 16 partition ( right )

Motion vectors of the current block are usually correlated with the motion vectors
of neighboring blocks in the current picture or in the earlier coded pictures. This is
because neighboring blocks are likely to correspond to the same moving object with
similar motion and the motion of the object is not likely to change abruptly over
time. Consequently, using the motion vectors in neighboring blocks as predictors
reduces the size of the signaled motion vector difference. The MVPs are usually
derived from already decoded motion vectors from spatial neighboring blocks or
from temporally neighboring blocks in the co-located picture. 1 In H.264/AVC, this
is done by doing a component wise median of three spatially neighboring motion
vectors. Using this approach, no signaling of the predictor is required. Temporal
MVPs from a co-located picture are only considered in the so called temporal direct
mode of H.264/AVC. The H.264/AVC direct modes are also used to derive other
motion data than the motion vectors. Hence, they relate more to the block merging
concept in HEVC and are further discussed in Sect. 5.2.2 .
In HEVC, the approach of implicitly deriving the MVP was replaced by a
technique known as motion vector competition, which explicitly signals which
MVP from a list of MVPs, is used for motion vector derivation [ 19 ]. The variable
coding quadtree block structure in HEVC can result in one block having several
neighboring blocks with motion vectors as potential MVP candidates. Taking the
left neighbor as an example, in the worst case a 64 64 luma prediction block
couldhave168 4 luma prediction blocks to the left when a 64 64 luma coding
tree block is not further split and the left one is split to the maximum depth.
Figure 5.3 illustrates this example and compares it to the worst case in H.264/AVC.
Advanced Motion Vector Prediction (AMVP) was introduced to modify motion
vector competition to account for such a flexible block structure [ 11 ]. During

1 In some cases, the zero motion vector can also be used as MVP.