Guided Color Grading for Extended Dynamic Range Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for color grading images and videos struggle to efficiently convert standard dynamic range (SDR) content into visual dynamic range (VDR) content, requiring extensive time and effort, especially when existing SDR versions need to be upgraded to VDR, and often result in loss of image quality due to differences in dynamic range and color gamut.
Innovation Solution
A system and method that uses a computer to perform automatic color grading by aligning SDR and ungraded images spatially and temporally, applying inverse tone mapping and gamut transformations, and estimating global and local color transformations to generate a higher dynamic range image, allowing for special handling regions and director-approved fine-tuning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual color grading methods are used to convert SDR content to VDR content, then color accuracy can be maintained, but the process requires extensive time and effort
Solution Approach 1:
The patent uses an inverse tone mapping operation as an intermediary step to transform the SDR graded image into a virtual HDR space. This intermediary transformation enables automatic color grading by providing a bridge between SDR and HDR domains, allowing algorithms to operate in the HDR space while leveraging existing SDR graded content as guidance.
Solution Approach 2:
The system changes the dynamic range parameter by applying inverse tone mapping to convert SDR values to HDR values. This parameter transformation allows the color grading process to automatically adapt to HDR requirements while maintaining the artistic intent from the SDR graded version, significantly reducing manual intervention time.
2Productivity
If existing SDR versions are upgraded to VDR, then content can be reused, but image quality is often lost due to differences in dynamic range and color gamut
Solution Approach 1:
The patent applies preliminary inverse tone mapping and color space transformations to the SDR graded content before using it to guide HDR color grading. This preliminary processing prepares the SDR content in a way that preserves its quality characteristics while adapting it to HDR requirements, enabling high-quality content reuse without manual re-grading.
Solution Approach 2:
The system creates a virtual copy of the SDR graded image in HDR space through inverse tone mapping. This copied representation maintains the original artistic intent and quality while existing in the HDR domain, allowing direct reuse of SDR graded content for HDR output without quality loss.
3Productivity
If automatic color grading algorithms are implemented, then processing speed increases, but accuracy in maintaining director's intentions decreases
Solution Approach 1:
The patent uses the SDR graded image as feedback to guide the automatic HDR color grading process. By comparing the inverse-mapped SDR graded content with the ungraded HDR content, the system continuously adjusts the color grading parameters to match the director's artistic intent captured in the SDR version, ensuring accuracy while maintaining automated processing speed.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
Novel methods and systems for color grading are disclosed. The color grading process for a visual dynamic range image can be guided by information relating to the color grading of other images such as the standard dynamic range image.