Tone Mapping Metadata for Local Dimming Displays

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Images encoded with higher bit depths often result in larger file sizes and increased computational requirements, leading to undesirable visually perceptible artifacts when displayed on lower bit depth displays, such as banding, especially in regions with slow luminance variations.

Innovation Solution

A method and apparatus for tone mapping image data from a higher bit depth to a lower bit depth format, using metadata to redefine the mapping between image data values and luminance values, allowing for improved image quality by reducing steps between adjacent luminance levels and utilizing display capabilities like local dimming to enhance dynamic range and contrast.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If images are encoded with higher bit depths to reduce visible artifacts, then image quality is improved, but file size increases and computational requirements increase

Engineering Contradiction:
Improveimage qualityVSAvoidfile size
Core Design Contradiction:
Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The patent applies parameter changes by dynamically adjusting the bit depth of image data based on the specific characteristics of different image regions. Instead of uniformly encoding the entire image at high bit depth, the system identifies regions requiring high precision (such as areas with gradual luminance transitions) and applies higher bit depth only to those regions, while using lower bit depth for regions where artifacts are less noticeable. This selective parameter adjustment reduces overall file size while maintaining image quality where it matters most.

Inventive Principle:
Principle #35Parameter changes

2Manufacturing precision

If images are encoded with higher bit depths to reduce visible artifacts, then image quality is improved, but computational processing time increases

Engineering Contradiction:
Improveimage qualityVSAvoidcomputational processing time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent implements segmentation by dividing the image into multiple regions with different luminance characteristics and processing each region with appropriate bit depth. The system identifies regions with slow luminance variations (prone to banding artifacts) and applies higher bit depth processing only to those specific segments, while using lower bit depth for other regions. This segmented approach significantly reduces computational processing time compared to uniformly processing the entire image at high bit depth, while still maintaining image quality in critical areas.

Inventive Principle:
Principle #1Segmentation

3Quantity of substance

If lower bit depth is used for distribution, then file size is reduced, but visible artifacts such as banding appear in displayed images

Engineering Contradiction:
Improvefile sizeVSAvoidimage quality
Core Design Contradiction:
Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent applies dynamics by enabling the display device to dynamically adjust local luminance levels using local dimming technology. The system receives image data encoded at lower bit depth and uses metadata or analysis of the image content to identify regions prone to banding artifacts. The display then dynamically adjusts the backlight or pixel luminance in those specific regions to compensate for the quantization artifacts, effectively eliminating visible banding while maintaining the benefits of lower bit depth encoding for distribution.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP4053786B1Display of high dynamic range images on a local dimming display using bit depth reduction mapping metadata
Publication Date: 2023.12.27 DOLBY LABORATORIES LICENSING CORP
  • EP4053786B1 patent drawingFigure 1~2
  • EP4053786B1 patent drawingFigure 3
  • EP4053786B1 patent drawingFigure 4~5

AI summary

Image data is encoded for distribution in a lower bit-depth format. The image data has a range that is less than a maximum range and is mapped to the lower bit depth format using a mapping such that a ratio of a range of the lower bit depth representation to a maximum range of the lower bit depth representation is greater than a ratio of the range of the image data to a maximum range of the image data. Metadata characterizing the mapping is associated with lower bit depth representation. The metadata may be used downstream in combination with local dimming capabilities of a display to reverse the mapping so that tonal detail is better reproduced.