Saturation Processing for HDR to LDR Dynamic Range Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image and video coding technologies struggle to effectively handle high dynamic range (HDR) images, particularly in converting and encoding images with significantly different luminance dynamic ranges, as they lack a simple and efficient saturation specification strategy that can handle large dynamic range conversions while maintaining color accuracy and avoiding clipping issues.
Innovation Solution
A color saturation modification apparatus that determines linear color differences and applies a gain based on a luminance-dependent function, allowing for saturation boost or reduction across the color gamut, specifically using a V-Y index to specify saturation modification, enabling efficient conversion between HDR and Low Dynamic Range (LDR) images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional saturation processing is used for HDR to LDR conversion, then color saturation can be adjusted, but clipping issues occur and color accuracy deteriorates
Solution Approach 1:
The patent changes the parameter space by introducing a new saturation specification method that operates in linear RGB space with luminance-dependent gain factors. Instead of using conventional saturation processing in non-linear color spaces, the invention applies gain factors gR, gG, gB to the linear color differences (R-Y), (G-Y), (B-Y) where the gains are determined as functions of luminance Y. This parameter transformation resolves the clipping issue by maintaining proper luminance-saturation relationships throughout the dynamic range conversion.
Solution Approach 2:
The patent segments the saturation processing by applying different gain factors to different color channels (R, G, B) based on their respective luminance contributions. The saturation specification function is divided into channel-specific gain functions gR(Y), gG(Y), gB(Y), each optimized for its color channel's characteristics. This segmentation allows precise control over color saturation in each channel while maintaining overall color accuracy during HDR to LDR conversion.
2Manufacturing precision
If saturation processing is applied to handle large dynamic range conversions, then color representation can be optimized, but processing complexity increases
Solution Approach 1:
The patent extracts the saturation control functionality from complex multi-dimensional color space transformations and isolates it into a simple luminance-dependent gain mechanism. By separating saturation processing from luminance mapping and applying it in linear RGB space, the invention reduces processing complexity while maintaining color representation accuracy. The gain factors gR, gG, gB are computed directly from luminance Y without requiring iterative optimization or complex lookup tables.
Solution Approach 2:
Instead of applying saturation processing in conventional non-linear color spaces (YUV, YCbCr) where clipping problems occur, the patent inverts the approach by working in linear RGB space where the relationship between luminance and color differences is straightforward. The saturation specification function is applied to linear color differences (R-Y), (G-Y), (B-Y) and then transformed back, which simplifies the mathematical operations and reduces processing complexity compared to traditional methods.
3Ease of manufacture
If conventional saturation processing is used, then implementation is straightforward, but it cannot handle HDR images with significantly different luminance dynamic ranges
Solution Approach 1:
The patent introduces dynamic saturation control where the gain factors gR, gG, gB are not fixed but vary as functions of luminance Y. This dynamic adaptation allows the saturation processing to automatically adjust to different luminance levels encountered in HDR images, providing versatility for handling large dynamic range conversions. The saturation specification function F(Y) enables real-time adaptation to local luminance conditions while maintaining a relatively simple implementation structure.
4Manufacturing precision
If saturation is boosted to enhance colorfulness, then perceived color quality improves, but clipping occurs in highlights
Solution Approach 1:
The patent changes the saturation control parameter from a fixed or uniform gain to luminance-dependent gain factors gR(Y), gG(Y), gB(Y). By making the saturation gain a function of luminance, the system can boost saturation in mid-tones and shadows where it enhances perceived color quality without causing clipping, while automatically reducing or maintaining appropriate saturation levels in highlights where clipping would occur. This parameter transformation elegantly resolves the trade-off between colorfulness and clipping.
Data Source
AI summary
Because we needed a new color saturation processing in tune with dynamic range transformations necessary for handling the recently introduced high dynamic range image encoding, we describe a color saturation modification apparatus (101) arranged to determine linear color differences (R-Y,G-Y,B-Y) on the basis of an input color (R,G,B) and a luminance (Y) of the input color, and to do a multiplication of the linear color differences (R-Y,G-Y,B-Y) with a gain (g), characterized in that the apparatus is arranged to determine the gain as a function of a difference value (V_in-Y) being defined as the value of the highest one of the linear color differences (R-Y,G-Y,B-Y).