FIELD: computer engineering.
SUBSTANCE: invention relates to computer engineering. Method of a video data decoding comprises receiving a bit stream by a video decoder, which comprises a network abstraction level (NAL) units sequence, wherein the NAL units sequence includes a non-level video encoding NAL unit (non-VCL); based on a non-VCL NAL unit having a nal_unit_type equal to EOS_NUT or EOB_NUT, or is in the range RSV_NVCL44…RSV_NVCL47 or in the range UNSPEC48…UNSPEC63, associating by video decoder of said the non-VCL NAL unit with the most recent NAL VCL unit preceding the non-VCL NAL unit in the decoding order, wherein the decoding unit (DU) consists of one or more VCL NAL units and NAL non-VCL units associated with one or more VCL NAL units, wherein one or more of NAL VCL units and non-VCL NAL units are associated with one or more VAL NAL units in the NAL units sequence; at a certain deletion time from the coded pictures buffer (CPB) of said DU, deletion by video decoder of the video data associated with the DU, from the CPB, and decoding of said video data associated with said DU.
EFFECT: technical result consists in increasing of errors tolerance.
8 cl, 16 dwg, 4 tbl
