Inter-view Motion Vector Derivation Using Depth Values

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current video signal processing techniques face challenges in achieving efficient coding of inter-view motion vectors, which are crucial for accurate inter-prediction in multiview video coding.

Innovation Solution

A method and device that obtain a reference view block using a predetermined fixed inter-view motion vector, derive an inter-view motion vector from depth values of reference blocks, and use this vector for decoding the current block, thereby improving prediction accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a predetermined fixed inter-view motion vector is used to obtain reference view block, then the device complexity is reduced, but the measurement precision of inter-prediction accuracy deteriorates

Engineering Contradiction:
Improvecomplexity of inter-view motion vector processingVSAvoidaccuracy of inter-prediction
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent segments the inter-view motion vector derivation process into two distinct phases: (1) initial acquisition using a predetermined fixed inter-view motion vector to reduce complexity, and (2) subsequent refinement using depth map information to improve precision. This segmentation allows the system to balance computational load with prediction accuracy by handling different processing stages separately.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies preliminary action by first obtaining the reference view block using a predetermined fixed inter-view motion vector before refining the motion vector using depth map data. This preliminary step ensures that a baseline prediction is available early in the decoding process, reducing initial computational complexity while allowing for later precision improvements.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If depth values of reference blocks are used to derive inter-view motion vector, then the measurement precision of inter-prediction is improved, but the loss of time for processing increases

Engineering Contradiction:
Improveaccuracy of inter-view motion vectorVSAvoidprocessing time for motion vector derivation
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies partial action by using only the necessary depth values from the depth map for motion vector derivation rather than processing all depth information. This selective approach improves the accuracy of the inter-view motion vector while minimizing the additional processing time required, as only relevant depth data is extracted and utilized for the specific purpose of motion compensation.

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If all depth samples within the associated depth block are used for disparity vector derivation, then the measurement precision is improved, but the device complexity increases

Engineering Contradiction:
Improveaccuracy of disparity vectorVSAvoidcomplexity of depth processing
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the essential depth values from the depth block that are necessary for accurate disparity vector derivation, rather than processing all depth samples. This extraction approach maintains measurement precision by selecting critical depth information while reducing device complexity by eliminating unnecessary processing of redundant depth data.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP2986002B1Video signal processing method and device
Publication Date: 2021.06.09 LG ELECTRONICS INC
  • EP2986002B1 patent drawingFigure 1
  • EP2986002B1 patent drawingFigure 2
  • EP2986002B1 patent drawingFigure 3

AI summary

The present invention relates to a video signal processing method and device capable of: obtaining a reference view block by using a predetermined motion vector; obtaining the depth value of a reference block which corresponds to the reference view block; obtaining an inter-view motion vector for a current block by using at least one depth value of the reference depth block; and decoding the current block by using the inter-view motion vector.