FIELD: information technology.
SUBSTANCE: invention relates to encoding and decoding video. Method of decoding video comprises receiving a bit stream, including data on a picture, information on size of maximum coding unit and information of splitting coding unit. Picture is broken down into maximum coding units, which are hierarchically broken down into coding units. Method includes determining prediction units and transformation units of coding units. Prediction is performed on prediction unit and inverse transformation on transformation unit is performed in order to generate a recovered coding unit. If boundary corresponds to boundary of prediction unit or boundary of transformation unit, power of boundary is determined based on nonzero transformation coefficients, prediction mode, motion vector or reference index. Method for deblocking filtration is determined and deblocking filtration is carried out in order to generate a filtered coding unit.
EFFECT: improved quality of compressed and recovered video.
1 cl, 27 dwg, 1 tbl
