FIELD: information technology.
SUBSTANCE: invention relates to encoding and decoding video. Video motion compensation method comprises receiving a partition type of a current coding unit and determining a size of a current prediction unit. Reference image L0 list, comprising a direct reference image and a reverse reference image, and L1 list comprising a reverse reference image and a direct reference image are being determined. Bit is being obtained, said bit indicates one of an L0 prediction, and an L1 prediction and bi-prediction for the current prediction unit by performing context, based entropy decoding on the bitstream using the first or the second context model for the first or the second element of the list, and L0 prediction, L1 prediction or bi-prediction are being performed according to the bit. L0 prediction accesses the L0 list, L1 prediction accesses the L1 list and bi-prediction accesses the L0 list and the L1 list.
EFFECT: technical results are reduced number of transmitted bits and simplified syntax data analysis.
3 cl, 27 dwg, 1 tbl
