FIELD: physics.
SUBSTANCE: invention relates to encoding and decoding video. Technical result is achieved by generating information of subdivision of coding unit of current depth into maximum unit of image coding, if the current depth encoding unit is not broken down into the lower depth encoding units, the current depth unit transmitting information is generated, if unit of encoding of current depth is divided into units of encoding of lower depth, coding unit of current depth is divided independently from neighbouring coding units into coding units of lower depth, and if the unit of encoding of the current depth is not broken into lower depth encoding units, at least one segment is determined from the unit of encoding of the current depth and prediction is performed for the unit of encoding of the current depth on the basis of at least one segment.
EFFECT: technical result is higher accuracy of transmitting encoded data.
11 cl, 23 dwg, 3 tbl
