FIELD: information technologies.
SUBSTANCE: invention relates to video encoding/decoding technologies. A method of decoding video is disclosed. Method includes a step, on which conversion flag is obtained from bit-stream with division for current depth, upon that, when conversion flag with division indicates division absence for current depth, determining, that conversion depth is equal to current depth. Further, according to method determining context increment parameter for context index determining based on fact, whether conversion depth is equal to certain predetermined value, without using conversion block size. Obtaining context model for conversion unit significant coefficient flag from among plurality of context models using context index, obtained by context increment and context displacement parameter addition.
EFFECT: technical result is simplification of entropy encoding/decoding operation due to context model selection based on conversion depth.
3 cl, 27 dwg, 1 tbl
