FIELD: encoding.
SUBSTANCE: invention is related in particular to performing inter-prediction based on decoder-side motion vector refinement (DMVR) and/or bidirectional optical flow (BDOF). A method for decoding images is proposed, which contains: extracting L0- and L1-vectors of movement of the current block; extracting detailed L0 and L1 motion vectors by applying the DMVR to the current block based on the DMVR flag information indicating whether or not to apply the DMVR to the current block; extracting predictive L0 samples of the current block based on the detailed L0 motion vector and predictive L1 samples of the current block based on the detailed L1 motion vector; extracting predicted samples of the current block based on the predicted L0 and L1 samples by applying BDOF to the current block based on BDOF flag information indicating whether or not to apply BDOF to the current block; and generating reconstruction samples of the current block based on the prediction samples, wherein the DMVR flag information is called based on the preset DMVR application condition, and the BDOF flag information is called out based on the preset BDOF application condition.
EFFECT: increase of efficiency of encoding and decoding of high-quality images and high-resolution videos having various characteristics.
15 cl, 22 dwg, 42 tbl
