FIELD: physics.
SUBSTANCE: invention relates to image processing. Method comprises analysis of an enable flag to refine a decoder motion vector (DMVR) from a set of video bitstream sequence parameters, wherein the DMVR enabling flag indicates whether DMVR-based inter-prediction is enabled for the plurality of pictures associated with the sequence parameter set, analysis of the disconnection flag for the DMVR from the syntax structure of a lower hierarchical level of the syntax structure than the sequence parameter set (SPS) level of the video bitstream, wherein the disable flag for the DMVR indicates whether the DMVR-based inter-prediction is disabled for at least one region of the current snapshot associated with the syntax, wherein the current picture belongs to the plurality of pictures, and performing the DMVR-based inter-prediction for the current block within the at least one region of the current picture to obtain the predicted values of the discrete samples of the current block when at least one predetermined condition is met, and wherein the at least one predetermined condition comprises that the analyzed value of the shutdown flag indicates that the DMVR-based inter-prediction is enabled for at least one region of the current snapshot.
EFFECT: high coding efficiency.
47 cl, 18 dwg
