Viewpoint-Specific Color Grading for Interactive 3D Content

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current colour grading methods are inadequate for interactive content such as video games and 3D environments, as they fail to provide a detailed and dynamic treatment of colour and tone mapping due to the varied viewpoints and experiences of players, which are not effectively addressed by existing methods based on lookup tables and frame-by-frame processing.

Innovation Solution

The solution involves using virtual cameras to capture and store multiple viewpoints of a scene with associated metadata, generating a colour data graph to identify consistent and varying lighting conditions, and applying colour grading and tone mapping operations on a per-viewpoint basis, either by adjusting light sources or modifying the scene's colour properties, allowing for real-time rendering of desired effects.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If a single lookup table is used for colour grading across all frames, then processing speed is improved, but colour grading accuracy and consistency across different viewpoints deteriorates

Engineering Contradiction:
Improveprocessing speedVSAvoidcolour grading accuracy
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The patent segments the single lookup table approach into multiple viewpoint-specific lookup tables. Each viewpoint has its own lookup table that stores pre-calculated colour grading and tone mapping parameters for that specific camera angle. This segmentation allows the system to quickly retrieve the appropriate lookup table based on the current viewpoint while maintaining accurate, consistent colour grading tailored to each perspective, thereby resolving the contradiction between processing speed and colour grading accuracy.

Inventive Principle:
Principle #1Segmentation

2Stability of the object's composition

If colour grading is applied to all elements in a scene, then visual consistency is improved, but processing time and computational resources increase

Engineering Contradiction:
Improvevisual consistencyVSAvoidprocessing time
Core Design Contradiction:
Stability of the object's compositionVSLoss of time

Solution Approach 1:

The patent implements local quality by selectively applying colour grading and tone mapping only to relevant elements in the scene based on the current viewpoint. The system identifies which objects or regions are visible and important from the player's perspective, then applies the appropriate lookup table parameters only to those elements. This approach maintains visual consistency for visible elements while significantly reducing processing time and computational resources by excluding invisible or irrelevant elements from processing.

Inventive Principle:
Principle #3Local quality

3Manufacturing precision

If detailed per-viewpoint colour grading is implemented, then colour grading accuracy is improved, but device complexity increases

Engineering Contradiction:
Improvecolour grading accuracyVSAvoidsystem complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing colour grading and tone mapping parameters in viewpoint-specific lookup tables during an offline preparation phase. Each lookup table contains pre-computed parameters for different lighting conditions, times of day, and environmental factors specific to each viewpoint. During runtime, the system simply retrieves the appropriate pre-computed parameters based on the current viewpoint and conditions, avoiding the need for complex real-time calculations and thereby reducing device complexity while maintaining high colour grading accuracy.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS11037363B2Content modification device and method
Publication Date: 2021.06.15 SONY INTERACTIVE ENTERTAINMENT LLC
  • US11037363B2 patent drawing
  • US11037363B2 patent drawing
  • US11037363B2 patent drawing

AI summary

A content modification system for modifying computer-generated content, the system including a rendering unit operable to render an image for display in dependence upon content to be rendered, an information output unit operable to output viewpoint data for the rendered images, the viewpoint data comprising information enabling the viewpoint to be reconstructed at a later time, a viewpoint reconstruction unit operable to generate an image to be displayed using the outputted viewpoint data, an image modifying unit operable to modify the generated image; and a content modifying unit operable to modify the content to be rendered in dependence upon the modifications to the generated image.