Multi-view Video Illumination and Color Compensation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Multi-view Video Coding systems face inefficiencies due to luminance and chrominance differences between heterogeneous cameras, which affect cross-view prediction, and existing methods either focus on temporal sequences or have limitations in handling significant illumination mismatches.

Innovation Solution

The implementation of illumination and color compensation methods within the Multi-view Video Coding framework, using new syntax elements like ic_prediction_flag, ic_enable, and ic_sym for illumination compensation, and extending these to color compensation, allows for adaptive block-level and slice-level adjustments to address illumination and color discrepancies across different camera views.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If illumination compensation is applied using block-by-block parameters, then prediction accuracy improves, but coding overhead increases

Engineering Contradiction:
Improveprediction accuracyVSAvoidcoding overhead
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent divides the picture into slices and further into blocks, applying illumination compensation parameters at different granularities. Slice-level parameters provide coarse correction while block-level parameters provide fine correction, segmenting the compensation task to balance accuracy and overhead.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different illumination compensation strategies to different regions: slice-level parameters for uniform regions and block-level parameters for regions with significant illumination variation. This local quality approach ensures high prediction accuracy where needed while minimizing overhead in uniform areas.

Inventive Principle:
Principle #3Local quality

2Productivity

If color compensation is applied to correct chrominance differences, then cross-view prediction efficiency improves, but device complexity increases

Engineering Contradiction:
Improvecross-view prediction efficiencyVSAvoiddevice complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent merges illumination compensation and color compensation into a unified cross-view prediction framework. By combining these functions and reusing the same block matching and parameter coding infrastructure, the patent improves prediction efficiency while avoiding the need for separate complex processing paths.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent makes the illumination compensation mechanism universal by extending it to handle both luminance and chrominance components. The same syntax elements and processing steps are used for both Y and UV channels, reducing device complexity through multi-functionality rather than separate dedicated systems.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Device complexity

If global illumination compensation is used, then device complexity is reduced, but prediction accuracy deteriorates when illumination mismatches are significant

Engineering Contradiction:
Improvedevice complexityVSAvoidprediction accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent implements a dynamic illumination compensation system that adapts to local conditions. The ic_enable flag and ic_sym parameters allow the system to switch between global and local compensation modes dynamically, ensuring high prediction accuracy in regions with significant illumination mismatches while maintaining low complexity in uniform regions.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP2424245B1Methods and apparatus for illumination and color compensation for multi-view video coding
Publication Date: 2019.09.04 DOLBY INTERNATIONAL AB
  • EP2424245B1 patent drawingFigure 1
  • EP2424245B1 patent drawingFigure 2
  • EP2424245B1 patent drawingFigure 3

AI summary

There are provided a method and apparatus for illumination and color compensation for multi-view video coding. A video encoder includes an encoder (100) for encoding a picture by enabling color compensation of at least one color component in a prediction of the picture based upon a correlation factor relating to color data between the picture and another picture. The picture and the other picture have different view points and both corresponding to multi-view content for a same or similar scene.