Back to results
US 12,764,320 B2Granted
Method and device for performing motion compensation using merge candidate list
AssigneeKt Corporation
InventorBae Keun Lee
Life of this patent
Estimating the term…
Abstract
Provided are a method and device for decoding image using merge candidate list. The method may include: deriving a spatial merge candidate of a current block; generating a merge candidate list for the current block on the basis of the spatial merge candidate; acquiring motion information on the current block on the basis of the merge candidate list; and performing motion compensation for the current block by using the motion information.
Claims4
A method of decoding a video, the method comprising:
- determining whether a merge mode is applied to a coding block or not;
- in response to the merge mode being applied to the coding block, deriving spatial merge candidates from spatial neighboring blocks, the spatial neighboring blocks being determined based on a position and a size of the coding block;
- constructing a merge candidate list of the coding block;
- obtaining motion information based on a merge index of the coding block;
- and
- performing motion compensation based on the motion information,
- wherein if a number of merge candidates included in the merge candidate list is less than a first threshold value, an average merge candidate, generated based on two of the merge candidates, is newly inserted to the merge candidate list, the merge candidates including the spatial merge candidates and a temporal merge candidate,
- wherein when a first merge candidate, among the two of the merge candidates, has both L0 (List 0) and L1 (List 1) motion information while a second merge candidate, among the two of the merge candidates, has only L0 motion information, an L0 motion vector of the average merge candidate is derived by averaging an L0 motion vector of the first merge candidate and an L0 motion vector of the second merge candidate and an L1 motion vector of the average merge candidate is derived to be the same as an L1 motion vector of the first merge candidate, and
- wherein an L0 reference picture index of the average merge candidate is inherited from one of the first merge candidate and the second merge candidate that has an index value less than the other of the first merge candidate and the second merge candidate.
6 more elements
A method of encoding a video, the method comprising:
- in response to a merge mode being applied to a coding block, deriving spatial merge candidates from spatial neighboring blocks, the spatial neighboring blocks being determined based on a position and a size of the coding block;
- constructing a merge candidate list of the coding block;
- obtaining motion information based on a index of the coding block;
- and
- performing motion compensation based on the motion information,
- wherein if a number of merge candidates included in the merge candidate list is less than a first threshold value, an average merge candidate, generated based on two of the merge candidates, is newly inserted to the merge candidate list, the merge candidates including the spatial merge candidates and a temporal merge candidate,
- wherein when a first merge candidate, among the two of the merge candidates, has both L0 (List 0) and L1 (List 1) motion information while a second merge candidate, among the two of the merge candidates, has only L0 motion information, an L0 motion vector of the average merge candidate is derived by averaging an L0 motion vector of the first merge candidate and an L0 motion vector of the second merge candidate and an L1 motion vector of the average merge candidate is derived to be the same as an L1 motion vector of the first merge candidate, and
- wherein an L0 reference picture index of the average merge candidate is inherited from one of the first merge candidate and the second merge candidate that has an index value less than the other of the first merge candidate and the second merge candidate.
5 more elements
A device for transmitting compressed data associated with a video signal, comprising:
- a processor configured to obtain the compressed data;
- and
- a transmitting unit configured to transmit the compressed data,
- wherein obtaining the compressed data comprises:
- in response to a merge mode being applied to a coding block, deriving spatial merge candidates from spatial neighboring blocks, the spatial neighboring blocks being determined based on a position and a size of the coding block;
- constructing a merge candidate list of the coding block;
- obtaining motion information based on a index of the coding block;
- and
- performing motion compensation based on the motion information,
- wherein if a number of merge candidates included in the merge candidate list is less than a first threshold value, an average merge candidate, generated based on two of the merge candidates, is newly inserted to the merge candidate list, the merge candidates including the spatial merge candidates and a temporal merge candidates,
- wherein when a first merge candidate, among the two of the merge candidates, has both L0 (List 0) and L1 (List 1) motion information while a second merge candidate, among the two of the merge candidates, has only L0 motion information, an L0 motion vector of the average merge candidate is derived by averaging an L0 motion vector of the first merge candidate and an L0 motion vector of the second merge candidate and an L1 motion vector of the average merge candidate is derived to be the same as an L1 motion vector of the first merge candidate, and
- wherein an L0 reference picture index of the average merge candidate is inherited from one of the first merge candidate and the second merge candidate that has an index value less than the other of the first merge candidate and the second merge candidate.
9 more elements
Dependencies are inferred from the claim text.