Granted Patent B2

US 12,695,863 · App. 18/959,930 · Granted Jul 28, 2026

Inventors: Joo Hee Moon (Seoul, KR); Dong Jae Won (Goyang-si, KR); Sung Won Lim (Seoul, KR)

Assignee: INDUSTRY ACADEMY COOPERATION FOUNDATION OF SEJONG UNIVERSITY

H04N19/103H04N19/117H04N19/119H04N19/159H04N19/176H04N19/51

Abstract

Claims (37)

1 . An image decoding method comprising:

determining whether a hybrid mode is applied to a current block as a prediction mode of the current block; and

generating, in response to the hybrid mode being applied to the current block, a first prediction sample of a first sub-block and a second prediction sample of a second sub-block,

wherein the current block is divided into two sub-blocks including the first sub-block and the second sub-block based on a division line, the division line connecting a right boundary of the current block and bottom boundary of the current block,

wherein the second sub-block is smaller than the first sub-block,

wherein whether the hybrid mode is applied to the current block is determined in a unit of the current block,

wherein first prediction information used to generate the first prediction sample includes index information specifying one candidate of a merge candidate list,

wherein the index information is obtained from a bitstream,

wherein second prediction information used to generate the second prediction sample is derived based on the first prediction information,

wherein the merge candidate list is determined based on a spatial neighboring block and a temporal collocated block of the current block,

wherein the spatial neighboring block is located on a neighboring location of the current block, and

wherein the neighboring location is adjacent to top-left sample, top-right, or bottom-left sample of the current block.

2 . An image encoding method comprising:

determining whether a hybrid mode is applied to a current block as a prediction mode of the current block; and

generating, in response to the hybrid mode being applied to the current block, a first prediction sample of a first sub-block and a second prediction sample of a second sub-block,

wherein the current block is divided into two sub-blocks including the first sub-block and the second sub-block based on a division line, the division line connecting a right boundary of the current block and bottom boundary of the current block,

wherein the second sub-block is smaller than the first sub-block,

wherein whether the hybrid mode is applied to the current block is determined in a unit of the current block,

wherein first prediction information used to generate the first prediction sample includes index information specifying one candidate of a merge candidate list,

wherein the index information is encoded into a bitstream,

wherein second prediction information used to generate the second prediction sample is derived based on the first prediction information,

wherein the merge candidate list is determined based on a spatial neighboring block and a temporal collocated block of the current block,

wherein the spatial neighboring block is located on a neighboring location of the current block, and

wherein the neighboring location is adjacent to top-left sample, top-right, or bottom-left sample of the current block.

3 . A method of transmitting a bitstream generated by an encoding method, the encoding method comprising:

determining whether a hybrid mode is applied to a current block as a prediction mode of the current block;

generating, in response to the hybrid mode being applied to the current block, a first prediction sample of a first sub-block and a second prediction sample of a second sub-block; and

transmitting the bitstream obtained based on the first prediction sample and the second prediction sample,

wherein the current block is divided into two sub-blocks including the first sub-block and the second sub-block based on a division line, the division line connecting a right boundary of the current block and bottom boundary of the current block,

wherein the second sub-block is smaller than the first sub-block,

wherein whether the hybrid mode is applied to the current block is determined in a unit of the current block,

wherein first prediction information used to generate the first prediction sample includes index information specifying one candidate of a merge candidate list,

wherein the index information is encoded into the bitstream,

wherein second prediction information used to generate the second prediction sample is derived based on the first prediction information,

wherein the merge candidate list is determined based on spatial neighboring blocks and a temporal collocated block of the current block,

wherein the spatial neighboring block is located on a neighboring location of the current block, and

wherein the neighboring location is adjacent to top-left sample, top-right, or bottom-left sample of the current block.

Priority Claims (3)

Continuity (6)

References Cited (110)