Method and apparatus for encoding/decoding motion vector information
AssigneesSamsung Electronics Co., Ltd., Sungkyunkwan University Foundation for Corporate Collaboration
InventorsKwang-Pyo Choi, Yun-Je Oh, Young-Hun Joo, Byeung-Woo Jeon, Bong-Soo Jung
Life of this patent
Estimating the term…
Abstract
A method and apparatus for encoding/decoding motion vector information are provided, in which a coding unit is divided into a plurality of subunits, PMVs of the subunits are calculated, it is determined whether the PMVs satisfy a predetermined condition, and coded motion vector data are generated, which include information indicating an omission of the motion vectors of the subunits, if the PMVs satisfy the predetermined condition.
Claims10
A method for encoding motion vector information of a coding unit, comprising:
- dividing the coding unit into a plurality of subunits and calculating Predictive Motion Vectors (PMVs) of the plurality of subunits in a plurality of Motion Vector Predictive Modes (MVPMs);
- recovering video information of the coding unit by applying the PMVs to a plurality of reference pictures;
- calculating residual data between the recovered video information and original video information of the coding unit;
- encoding and decoding the residual data;
- generating re-recovered video information by applying the encoded and decoded residual data to the recovered video information;
- calculating a cost between the re-recovered video information and the original video information;
- selecting an MVPM with a minimum cost;
- setting the minimum cost as a first cost;
- recovering the video information of the coding unit by applying the PMVs to a single reference picture;
- calculating a second cost between the recovered video information and the original video information;
- determining if the PMVs satisfy a predetermined condition;
- if the PMVs satisfy the predetermined condition, generating a first coded bit stream comprising only a Motion Vector Coding Mode (MVCM) Identifier (ID), indicating that motion vectors of the plurality of subunits and information about the single reference picture are omitted in the first coded bit stream;
- if the PMVs do not satisfy the predetermined condition, generating a second coded bit stream comprising an MVCM ID, indicating that the motion vectors of the plurality of subunits are included in the second coded bit stream, information about the plurality of reference pictures, and the motion vectors of the plurality of subunits;
- and
- transmitting the first coded bit stream or the second coded bit stream to a video decoding apparatus,
- wherein the predetermined condition is that the second cost is less than the first cost and the PMVs of the plurality of subunits are equal to the motion vectors of the plurality of subunits.
An apparatus for encoding motion vector information about a coding unit using a motion estimator and a motion compensator, the apparatus comprising:
- means for recovering video information of the coding unit by applying Predictive Motion Vectors (PMVs) to a plurality of reference pictures;
- means for calculating residual data between the recovered video information and original video information of the coding unit, and encoding and decoding the residual data;
- means for generating re-recovered video information by applying the encoded and decoded residual data to the recovered video information;
- means for calculating a cost between the re-recovered video information and the original video information;
- means for selecting a Motion Vector Predictive Mode (MVPM) with a minimum cost and setting the minimum cost as a first cost;
- means for recovering the video information of the coding unit by applying the PMVs to a single reference picture and calculating a second cost between the recovered video information and the original video information;
- means for determining if the PMVs satisfy a predetermined condition;
- means for generating a first coded bit stream including only a Motion Vector Coding Mode (MVCM) Identifier (ID) indicating that motion vectors of the plurality of subunits and information about the single reference picture are omitted from the first coded bit stream, if the PMVs satisfy the predetermined condition;
- means for generating a second coded bit stream including an MVCM ID indicating that the motion vectors of the plurality of subunits are included in the second coded bit stream, information about the plurality of reference pictures, and the motion vectors of the plurality of subunits, if the PMVs do not satisfy the predetermined condition;
- and
- means for transmitting the first coded bit stream or the second coded bit stream to a video decoding apparatus,
- wherein the predetermined condition is that the second cost is less than the first cost and the PMVs of the plurality of subunits are equal to the motion vectors of the plurality of subunits.
Dependencies are inferred from the claim text.