Temporal Direct Motion Vector Selection for Video Coding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional temporal direct motion vector selection in moving picture coding is limited to forward reference pictures, making it difficult to derive the most suitable motion vector, which decreases the compression rate.

Innovation Solution

A method for adaptively selecting a motion vector in temporal direct coding by generating multiple predictive motion vectors and using them to code the current block, allowing for the selection of the most suitable vector based on reference motion vectors from co-located blocks, even if they are backward or have different directions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If conventional temporal direct motion vector selection is used with forward reference pictures only, then the coding process is simple, but the compression rate decreases

Engineering Contradiction:

Improvesimplicity of motion vector selectionVSAvoidcompression rate

Core Design Contradiction:

Ease of manufactureVSProductivity

Solution Approach 1:

The patent changes the parameter of reference picture selection by extending temporal direct mode to use backward reference pictures in addition to forward reference pictures. This allows the motion vector prediction to access a broader range of reference frames, improving compression efficiency while maintaining algorithmic simplicity

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent makes the temporal direct mode universally applicable to both forward and backward reference pictures. The motion vector prediction mechanism can now function with any reference picture type, enhancing its versatility and effectiveness in various coding scenarios

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Device complexity

If motion vector selection is limited to forward reference pictures, then the reference picture list management is simple, but the coding efficiency is reduced

Engineering Contradiction:

Improvereference picture list management complexityVSAvoidcoding efficiency

Core Design Contradiction:

Device complexityVSLoss of information

Solution Approach 1:

The patent modifies the reference picture list usage parameter by enabling temporal direct mode to select from both forward and backward reference pictures. This parameter change allows more accurate motion vector prediction without significantly increasing list management complexity

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent performs preliminary selection of the most suitable reference picture from the available lists before motion compensation. By pre-identifying the optimal reference frame based on temporal proximity and direction, the system achieves better coding efficiency without complex real-time adjustments

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11284102B2Moving picture coding method, moving picture decoding method, moving picture coding apparatus, moving picture decoding apparatus, and moving picture coding and decoding apparatus

Publication Date: 2022.03.22 SUN PATENT TRUST

AI summary

A moving picture coding method includes: coding, using a motion vector, a current block to be coded; generating a plurality of predictive motion vectors; and coding the motion vector using one of the predictive motion vectors, and when a co-located block included in a coded picture and co-located with the current block has two reference motion vectors and reference directions of the two reference motion vectors are the same, a first prediction vector is generated using a first reference motion vector and a second prediction vector is generated using a second reference motion vector in the generating of a plurality of predictive motion vectors.