Image encoding/decoding method and apparatus

Inventor: Ki Baek Kim (Daejeon, KR)

Assignee: B1 INSTITUTE OF IMAGE TECHNOLOGY, INC

H04N19/51 H04N19/105 H04N19/159 H04N19/176

View Patent ↗

Loading inventors, assignments & file history…

Monitor This Case

Get email alerts when status or documents change.

Order Certified Copies

Most orders are placed with the USPTO same day — all within 24 business hours.

Order via The Patent Place →

Pre-filled with this patent's details

Quick Facts

Patent No.: US 11,758,175
App. No.: 17/252,019
Granted: Sep 12, 2023
Kind: B2

External Links

USPTO Patent Center↗ Full Patent PDF↗

Abstract

An image encoding/decoding method and apparatus according to the present invention may determine a motion information encoding mode of a target block, configure a motion information prediction candidate group according to the motion information encoding mode, select one candidate from the motion information prediction candidate group so as to induce the selected candidate as motion information of the target block, and perform inter-prediction on the target block on the basis of the motion information of the target block.

Claims (49)

1. An inter prediction method performed by an image decoding apparatus, comprising:

specifying a reference block of a target block according to an affine mode;

determining a first motion vector prediction candidate from motion vectors of the reference block in case that the reference block is predicted by the affine mode;

determining a second motion vector prediction candidate by combining a plurality of translational motion vectors from more than one reference blocks;

determining a motion vector prediction candidate group sequentially comprising the first motion vector prediction candidate and the second motion vector prediction candidate;

deriving motion vectors of the target block for the affine mode from the motion vector prediction candidate group; and

performing inter prediction by using the motion vector of the target block for the affine mode,

wherein the deriving motion vector of the target block from the motion vector prediction candidate group comprises:

deriving a prediction motion vector of the target block from the motion vector prediction candidate group;

determining a difference motion vector precision among a plurality of precision candidates based on difference motion vector precision information, the difference motion vector precision only applied to a difference motion vector;

determining the difference motion vector of the target block;

scaling the difference motion vector of the target block based on a precision difference between the difference motion vector precision and a motion vector precision of the prediction motion vector, wherein the prediction motion vector of the target block is not scaled; and

deriving the motion vector of the target block based on the prediction motion vector of the target block and the scaled difference motion vector of the target block.

2. The method of claim 1 , wherein the affine mode is determined to be used for the target block by a predetermined first flag.

3. The method of claim 2 , wherein the first flag additionally derives a second flag which classifies a detailed mode.

4. The method of claim 3 , wherein the second flag is generated when motion information encoding mode is determined to be a skip mode or a merge mode.

5. The method of claim 1 ,

wherein the determining a motion vector precision among a plurality of precision candidates based on motion vector precision information comprises:

decoding adaptive motion vector precision enabled information indicating whether the motion vector precision is adaptively determined among the plurality of precision candidates or determined to be a default motion vector precision;

when the motion vector precision is determined to be adaptively determined among the plurality of precision candidates, decoding adaptive motion vector precision selection information indicating the motion vector precision among the plurality of precision candidates; and

determining the motion vector precision among the plurality of precision candidates based on motion vector precision selection information.

6. An inter prediction method performed by an image encoding apparatus, comprising:

deriving motion vectors of a target block for an affine mode;

performing inter prediction by using the motion vectors of the target block for the affine mode;

specifying a reference block of the target block according to the affine mode;

determining a first motion vector prediction candidate from motion vectors of the reference block in case that the reference block is predicted by the affine mode;

determining a second motion vector prediction candidate by combining a plurality of translational motion vectors from more than one reference blocks;

determining a motion vector prediction candidate group sequentially comprising the first motion vector prediction candidate and the second motion vector prediction candidate; and

encoding the motion vectors of the target block based on the motion vector prediction candidate group,

wherein the encoding motion vectors of the target block from the motion vector prediction candidate group comprises:

deriving a prediction motion vector of the target block from the motion vector prediction candidate group;

determining a difference motion vector of the target block based on the prediction motor vector of the target block;

determining a difference motion vector precision among a plurality of precision candidates;

encoding the motion vector of the target block based on the prediction motion vector of the target block and the scaled difference motion vector of the target block and encoding difference motion vector precision information indicating the difference motion vector precision.

7. A non-transitory computer readable recording medium storing a bitstream that is generated by an inter prediction method performed by an image encoding apparatus, the method comprising: