Scene-Referred Metadata for Accurate Color Mapping

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing image processing systems face inaccuracies when handling images from different cameras and settings due to limited color and gray level ranges, leading to undesirable visual effects in processed images.

Innovation Solution

The generation and use of scene-referred metadata, which includes reference color or gray levels and their corresponding coded values, allows for accurate processing of images by establishing correspondence relationships between reference and coded values, enabling more accurate image processing operations such as tone mapping and color grading.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If image processing operations are performed based on color values from limited image sampling, then processing speed is maintained, but measurement precision of color and gray levels deteriorates

Engineering Contradiction:

Improveprocessing speedVSAvoidcolor and gray level accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system performs preliminary action by capturing a reference image containing reference objects with known color and gray level values before processing the main scene image. This reference data is stored and used to establish correspondence relationships, enabling accurate color and gray level mapping without requiring extensive sampling during the actual processing phase, thus maintaining processing speed while improving measurement precision.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system introduces an intermediary reference image with known color and gray level values that mediates between the limited sampled color values and the required accurate color representation. By mapping sampled colors to the reference color space through correspondence relationships, the system achieves accurate color and gray level measurement without requiring exhaustive sampling of all possible color values.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If guesswork is used to compensate for missing colors and shades, then completeness of color range is improved, but manufacturing precision of color values deteriorates

Engineering Contradiction:

Improvecompleteness of color rangeVSAvoidaccuracy of color values

Core Design Contradiction:

Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The reference image serves as an intermediary that provides accurate color and gray level values for mapping. Instead of using guesswork to compensate for missing colors, the system establishes correspondence relationships between sampled colors and the comprehensive reference color space, enabling both complete color range coverage and high accuracy through the reference-mediated mapping process.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system changes the parameter representation by transforming sampled color values into the reference color space using established correspondence relationships. This parameter transformation allows the system to work with a complete color range defined by the reference image while maintaining accuracy through the mathematical mapping, avoiding the need for guesswork.

Inventive Principle:

Principle #35Parameter changes

3Device complexity

If images from different cameras and settings are processed without scene-referred metadata, then device complexity is reduced, but reliability of color consistency deteriorates

Engineering Contradiction:

Improvesimplicity of processing systemVSAvoidcolor consistency across images

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The system introduces scene-referred metadata as an intermediary that bridges images from different cameras and settings. By capturing reference images with each camera setup and establishing correspondence relationships, the metadata mediates the color and gray level values across different devices, ensuring consistent and reliable color representation while maintaining relatively simple processing through automated mapping.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The scene-referred metadata serves as a universal reference framework that works across multiple cameras and camera settings. By establishing correspondence relationships between each camera's color space and the reference color space, the system creates a multi-functional solution that ensures color consistency across diverse imaging devices without requiring complex device-specific processing for each camera.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS10114838B2Reference card for scene referred metadata capture

Publication Date: 2018.10.30 DOLBY LABORATORIES LICENSING CORP

AI summary

Scene-referred metadata comprising correspondence relationships between coded values used in generated images and reference values defined independent of any specific image may be provided as a part of image metadata for the generated images. Downstream image processing devices or image rendering devices may use the scene-referred metadata to perform image processing or rendering operations. When coded values of input images are altered in corresponding output images, the scene-referred metadata may be updated with new coded values used in the output images. Reference values refer to reference color values or reference gray levels. Coded values refer to color values or gray levels coded in pixels or sub-pixels of one or more images.