Image encoding/decoding method and device for performing PROF, and method for transmitting bitstream

    公开(公告)号:US11589053B2

    公开(公告)日:2023-02-21

    申请号:US17683203

    申请日:2022-02-28

    Abstract: An image encoding/decoding method and apparatus are provided. An image decoding method according to the present disclosure is performed by an image decoding apparatus. The image decoding method may comprise deriving a prediction sample of a current block based on motion information of the current block, determining whether prediction refinement with optical flow (PROF) applies to the current block, deriving, based on that the PROF applies to the current block, a difference motion vector for each sample position in the current block, deriving a gradient for each sample position in the current block, deriving a PROF offset based on the difference motion vector and the gradient, and deriving a refined prediction sample for the current block based on the PROF offset.

    IMAGE ENCODING/DECODING METHOD AND APPARATUS USING QUANTIZATION MATRIX, AND METHOD FOR TRANSMITTING BITSTREAM

    公开(公告)号:US20220345694A1

    公开(公告)日:2022-10-27

    申请号:US17762686

    申请日:2020-09-23

    Abstract: An image encoding/decoding method and apparatus are provided. An image decoding method performed by an image decoding apparatus comprises obtaining quantization mode information for a current block from a bitstream, determining whether a quantization matrix for the current block is constructed from a reference quantization matrix, based on the quantization mode information, and constructing the quantization matrix for the current block, based on a result of determination. When the quantization matrix for the current block is constructed from the reference quantization matrix, the quantization matrix for the current block may be constructed based on identification information of the reference quantization matrix. When the quantization matrix for the current block is not constructed from the reference quantization matrix, the quantization matrix for the current block may be constructed based on a value of the quantization matrix obtained from the bitstream. When the current block is an intra-predicted chroma block, the quantization matrix for the current block may be constructed only when a size of the current block is greater than 2×2.

Patent Agency Ranking