Piecewise-linear inter-layer predictor for HDR video coding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for coding and delivering high-dynamic range (HDR) images suffer from artifacts such as banding, blocking, and color bleeding, particularly in low bit rate scenarios due to the non-linear response of the human visual system and limitations in existing prediction techniques.

Innovation Solution

A piecewise-linear inter-layer predictor is employed to adjust the slope of the prediction function based on scene-significance histograms, masking residuals to improve encoding efficiency and reduce artifacts, while maintaining a low bit rate for the enhancement layer, and is compliant with existing Dolby Vision decoders.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If existing prediction techniques are used for HDR video coding, then encoding simplicity is maintained, but visual quality deteriorates due to artifacts such as banding, blocking, and color bleeding

Engineering Contradiction:
Improvevisual qualityVSAvoidcoding artifacts
Core Design Contradiction:
Manufacturing precisionVSObject-generated harmful factors

Solution Approach 1:

The prediction function is divided into multiple linear segments with different slopes, allowing different regions of the dynamic range to be predicted with different characteristics. This segmentation enables better adaptation to the non-linear response of the human visual system while avoiding the artifacts produced by uniform prediction approaches.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different segments of the prediction function are assigned different slopes tailored to specific luminance ranges. This local optimization allows the predictor to match the human visual system's non-linear response in different brightness regions, improving visual quality and reducing artifacts like banding and color bleeding in specific areas.

Inventive Principle:
Principle #3Local quality

2Manufacturing precision

If enhancement layer data is increased to improve visual quality, then HDR video quality improves, but bit rate increases

Engineering Contradiction:
Improvevisual qualityVSAvoidbit rate
Core Design Contradiction:
Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The prediction function parameters (slopes of different segments) are optimized to minimize the enhancement layer data required. By carefully selecting segment slopes that match the human visual system's characteristics, the residual error is reduced, allowing lower bit rates in the enhancement layer while maintaining high visual quality.

Inventive Principle:
Principle #35Parameter changes

3Manufacturing precision

If piecewise-linear prediction is implemented, then visual quality and artifact reduction are improved, but computational complexity increases

Engineering Contradiction:
Improvevisual qualityVSAvoidcomputational complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The piecewise-linear function divides the prediction into multiple segments, each with simple linear calculations. While this improves visual quality by adapting to different luminance ranges, it does increase computational complexity compared to a single linear predictor. The complexity is managed by limiting the number of segments and using efficient calculation methods.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3371975B1Piecewise-linear inter-layer predictor for high-dynamic range video coding
Publication Date: 2021.01.06 DOLBY LABORATORIES LICENSING CORP
  • EP3371975B1 patent drawingFigure 1~2
  • EP3371975B1 patent drawingFigure 3
  • EP3371975B1 patent drawingFigure 4~5

AI summary

In a method to improve the dynamic range of high-dynamic range (HDR) signals using an enhancement layer, a piecewise-linear inter-layer predictor and a residual masking operator are applied. The generation of the piecewise-linear inter-layer prediction function is based on a computed scene-significance histogram based on the average of frame-significance histograms indicating pixel values where coding artifacts are most likely to occur. For each segment in the prediction function, its slope is inversely proportional to a measure of energy in the segment under the scene-significance histogram. Bit rate constrains for the enhancement layer are also taken into consideration in determining the piecewise-linear prediction function.