Inter-layer residue prediction using pixel domain inverse tone mapping

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for inter-layer residue prediction in scalable video coding, particularly for bit-depth scalability, face challenges in applying inverse tone mapping in the residue domain, which hinders efficient decoding and compatibility with standards like MPEG-4 AVC.

Innovation Solution

Converting the inverse tone mapping problem from the residue domain to the pixel domain for inter-layer residue prediction, involving tone mapping of motion compensated predictions and upsampling of base layer residues, followed by filtering and inverse tone mapping in the spatial domain.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If inverse tone mapping is applied in the residue domain for inter-layer residue prediction, then bit-depth scalability is achieved, but decoding complexity increases and standard compatibility is compromised

Engineering Contradiction:
Improvebit-depth scalabilityVSAvoiddecoding complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

Instead of applying inverse tone mapping directly in the residue domain, the patent inverts the approach by: (1) applying tone mapping in the pixel domain to generate a tone-mapped picture, (2) deriving residues from this tone-mapped picture, and (3) using these derived residues for prediction. This inversion resolves the contradiction by maintaining bit-depth scalability while avoiding the complexity of direct inverse tone mapping in the residue domain.

Inventive Principle:
Principle #13The other way round (Inversion)

Solution Approach 2:

The patent introduces the tone-mapped picture as an intermediary between the original high-bit-depth picture and the residue prediction process. This intermediary allows the system to work with standardized low-bit-depth representations while preserving the ability to reconstruct high-bit-depth content, thus reducing decoding complexity while maintaining adaptability.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If tone mapping is used in pixel domain for inter-layer prediction, then standard compatibility is maintained, but inverse tone mapping in residue domain becomes difficult

Engineering Contradiction:
Improvestandard compatibilityVSAvoidinverse tone mapping difficulty
Core Design Contradiction:
ReliabilityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent inverts the traditional approach by performing tone mapping in the pixel domain first (maintaining standard compatibility) and then deriving residues from the tone-mapped picture. This avoids the difficult inverse tone mapping problem in the residue domain while preserving standard compatibility through the use of conventional tone mapping functions.

Inventive Principle:
Principle #13The other way round (Inversion)

3Measurement precision

If complete decoding including motion-compensated prediction and deblocking is performed for all layers, then prediction accuracy improves, but decoding complexity increases

Engineering Contradiction:
Improveprediction accuracyVSAvoiddecoding complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts and utilizes only the essential components needed for accurate prediction: motion-compensated prediction and deblocking filters. By selectively applying these specific processing steps rather than complete decoding of all layers, the system achieves good prediction accuracy while reducing unnecessary decoding complexity.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent performs motion-compensated prediction and deblocking as preliminary processing steps on the base layer before using its residues for enhancement layer prediction. This preliminary action prepares the data in advance, improving prediction accuracy while avoiding the need for complete redundant decoding of the enhancement layer.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2206349B1Methods and apparatus for inter-layer residue prediction for scalable video
Publication Date: 2015.03.11 THOMSON LICENSING SA
  • EP2206349B1 patent drawingFigure 1
  • EP2206349B1 patent drawingFigure 2
  • EP2206349B1 patent drawingFigure 3

AI summary

There are provided methods and apparatus for inter-layer residue prediction for scalable video. An apparatus is described for an encoder (200) for encoding a block of a picture, or a decoder (300) for decoding a block of a picture, by applying inverse tone mapping to an inter-layer residue prediction process for the block, wherein the inverse tone mapping is performed in the pixel domain. Methods for encoding (440, 460) or decoding (540, 560) a block of a picture are also described; and performed by applying inverse tone mapping to an inter-layer residue prediction process for the block, wherein the inverse tone mapping is performed in the pixel domain.