Motion Candidate List Construction for Video Coding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current moving-picture compression coding technologies, such as MPEG-4AVC/H.264, face inefficiencies in coding motion information, leading to larger code sizes and reduced prediction precision due to limitations in motion vector prediction methods.

Innovation Solution

A moving picture coding device and method that derive and merge motion information candidates from neighboring and temporal blocks, constructing a list of candidates to select the most efficient motion information for coding, thereby reducing code size and improving coding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the size of block used for motion compensation is reduced to improve prediction accuracy, then prediction precision is improved, but the code size of coding vector becomes large

Engineering Contradiction:

Improveprediction precisionVSAvoidcode size

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent merges spatial motion information from neighboring blocks with temporal motion information from reference blocks to create a combined motion candidate list. This merging allows the system to use larger block sizes for motion compensation while maintaining prediction accuracy through the combined spatial-temporal motion candidates, thereby reducing the code size required for transmitting motion vectors.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The motion candidate list structure is designed to serve multiple functions: it provides both spatial and temporal motion prediction candidates, supports different block sizes, and enables flexible selection based on coding conditions. This multi-functionality allows the system to achieve accurate motion compensation with larger blocks, reducing the number of coding vectors needed.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Device complexity

If only one prediction block and one direct mode are used to simplify the coding structure, then device complexity is reduced, but prediction precision deteriorates

Engineering Contradiction:

Improvecoding structure complexityVSAvoidprediction precision

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent implements a dynamic motion candidate list that adapts its content based on the coding context. The list is constructed by dynamically selecting and combining spatial and temporal motion candidates according to the specific block position, reference picture availability, and coding mode. This dynamic approach provides multiple prediction options without requiring a fixed complex structure, balancing simplicity and precision.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The motion prediction process is segmented into distinct stages: spatial candidate derivation, temporal candidate derivation, candidate list construction, and final selection. This segmentation allows each stage to be independently optimized and managed, maintaining overall system simplicity while achieving high prediction precision through the combination of multiple candidates.

Inventive Principle:

Principle #1Segmentation

3Quantity of substance

If temporal direct motion compensation prediction is used to reduce code size by scaling motion vectors, then code size is reduced, but prediction precision deteriorates due to loss of motion continuity in spatial direction

Engineering Contradiction:

Improvecode sizeVSAvoidprediction precision

Core Design Contradiction:

Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent combines spatial motion information from neighboring blocks with temporal motion information from reference blocks to create a unified motion candidate list. This combination preserves spatial motion continuity while still achieving code size reduction, as the system can select from multiple candidates including both spatial and temporal predictions without requiring transmission of full motion vectors for all cases.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The motion candidate list functions as a composite structure integrating both spatial and temporal motion information. Similar to composite materials combining different properties, this composite motion information provides the advantages of both spatial accuracy and temporal compression, allowing the system to achieve both reduced code size and maintained prediction precision.

Inventive Principle:

Principle #40Composite materials

Data Source

PatentUS9900596B2Moving picture coding device, moving picture coding method and moving picture coding program, and moving picture decoding device, moving picture decoding method and moving picture decoding program

Publication Date: 2018.02.20 JVC KENWOOD CORP

AI summary

A merging motion information candidate list construction unit generates a merging motion information candidate list, which is a list of merging motion information candidates, using a spatial merging motion information candidate and a temporal merging motion information candidate. A second merging motion information candidate supplying unit generates a new merging motion information candidate including, if a reference index indicates an available reference picture, the reference index and, if the reference index does not indicate an available reference picture, a predetermined reference index along with a motion vector having a size and a direction that are preset and adds the new merging motion information candidate to the merging motion information candidate list. A merging motion information selection unit selects one merging motion information candidate from the merging motion information candidate list to which the new merging motion information candidate has been added.