Image encoding/decoding method and device, and recording medium in which bitstream is stored for block division at picture boundary
In this specification, a image decoding method is disclosed. The image decoding method of the present invention comprises, decoding block partition information of a current block included in a current picture, determining a partitioning scheme for the current block according to the block partition information and partitioning the current block using the partitioning scheme determined, wherein the partitioning scheme is determined according to whether the current block includes a boundary of the current picture.
1 . A method of decoding an image, the method comprising:
dividing a current block of the image into a plurality of sub-blocks diagonally using triangular division mode; and
obtaining a prediction block of the current block based on reference blocks of the current block specified by a plurality of candidates in a candidate list of the current block, the reference blocks specified by the plurality of candidates being different each other, the candidate list being determined in a unit of the current block,
wherein obtaining the prediction block of the current block comprising:
obtaining, based on the reference blocks, motion information in sub-block units;
obtaining prediction blocks of sub-blocks through prediction in sub-blocks units based on the motion information; and
obtaining the prediction block of the current block by weighted summing the prediction blocks of the sub-blocks,
wherein the reference blocks are not determined in units of the sub-blocks, but are determined in units of the current block,
wherein the current block includes the diagonally divided sub-blocks in which the prediction block of the current block is obtained by weighted summing the prediction blocks of the sub-blocks,
wherein block information of the reference blocks of the current block includes L0 motion vector and L1 motion vector,
wherein a number of the plurality of candidates of the current block used to obtain the prediction block of the current block is 2, and
wherein two candidates used to obtain the prediction block of the current block are specified by two indexes signaled from a bitstream.
2 . The method of claim 1 , wherein the triangular division mode is used for dividing the current block only when a size of the current block is 8×8 or greater.
3 . The method of claim 1 , wherein the current block is divided based on division information signaled from the bitstream.
4 . A method of encoding an image, the method comprising:
dividing a current block of the image into a plurality of sub-blocks diagonally using triangular division mode; and
obtaining a prediction block of the current block based on reference blocks of the current block specified by a plurality of candidates in a candidate list of the current block, the reference blocks specified by the plurality of candidates being different each other, the candidate list being determined in a unit of the current block,
wherein obtaining the prediction block of the current block comprising:
obtaining, based on the reference blocks, motion information in sub-block units;
obtaining prediction blocks of sub-blocks through prediction in sub-blocks units based on the motion information; and
obtaining the prediction block of the current block by weighted summing the prediction blocks of the sub-blocks,
wherein the reference blocks are not determined in units of the sub-blocks, but are determined in units of the current block,
wherein the current block includes the diagonally divided sub-blocks in which the prediction block of the current block is obtained by weighted summing the prediction blocks of the sub-blocks,
wherein block information of the reference blocks of the current block includes L0 motion vector and L1 motion vector,
wherein a number of the plurality of candidates of the current block used to obtain the prediction block of the current block is 2, and
wherein two indexes specifying two candidates used to obtain the prediction block of the current block are encoded into a bitstream.
5 . The method of claim 4 , wherein the triangular division mode is used for dividing the current block only when a size of the current block is 8×8 or greater.
6 . The method of claim 4 , wherein division information determined based on division of the current block is encoded into the bitstream.
7 . A method of transmitting a bitstream, comprising:
dividing a current block of the image into a plurality of sub-blocks diagonally using triangular division mode;
obtaining a prediction block of the current block based on reference blocks of the current block specified by a plurality of candidates in a candidate list of the current block, the reference blocks specified by the plurality of candidates being different each other, the candidate list being determined in a unit of the current block; and
transmitting the bitstream generated based on the prediction block of the current block,
wherein obtaining the prediction block of the current block comprising:
obtaining, based on the reference blocks, motion information in sub-block units;
obtaining prediction blocks of sub-blocks through prediction in sub-blocks units based on the motion information; and
obtaining the prediction block of the current block by weighted summing the prediction blocks of the sub-blocks,
wherein the reference blocks are not determined in units of the sub-blocks, but are determined in units of the current block,
wherein the current block includes the diagonally divided sub-blocks in which the prediction block of the current block is obtained by weighted summing the prediction blocks of the sub-blocks,
wherein block information of the reference blocks of the current block includes L0 motion vector and L1 motion vector,
wherein a number of the plurality of candidates of the current block used to obtain the prediction block of the current block is 2, and
wherein two indexes specifying two candidates used to obtain the prediction block of the current block are encoded into the bitstream.