HDR Color Conversion Correction via Chroma Subsampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional HDR color conversion processes result in inaccuracies and artifacts due to chroma subsampling and differing upsampling filters, leading to reconstructed color values that differ significantly from the original, especially in regions with steep nonlinear transfer function slopes and near color gamut boundaries.
Innovation Solution
A method involving downsampling and upsampling of color space values, with the determination of new component values based on nonlinear transfer functions to correct color values, ensuring they fall within the color space range, and signaling the filter type to receiving devices for accurate reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If chroma subsampling is applied to reduce bitrate, then transmission efficiency is improved, but color accuracy deteriorates due to averaging artifacts in steep slope regions
Solution Approach 1:
The patent applies preliminary correction to the downsampled chroma values before upsampling. Specifically, it calculates correction values based on the difference between original and downsampled chroma values, then applies these corrections during the upsampling process. This preliminary action compensates for the averaging artifacts introduced by subsampling, particularly in regions with steep nonlinear transfer function slopes, thereby restoring color accuracy while maintaining the bitrate reduction benefits of subsampling.
2Device complexity
If conventional upsampling filters are used to reconstruct chroma values, then processing simplicity is improved, but reconstruction accuracy deteriorates due to filter mismatches
Solution Approach 1:
The patent implements a feedback mechanism where the system evaluates the impact of downsampling on chroma values and uses this information to adjust the upsampling process. By calculating correction values based on the original chroma values and the downsampled chroma values, the system provides feedback that compensates for the filtering effects. This feedback loop enables accurate reconstruction of chroma values without requiring complex inverse filters, thus maintaining processing simplicity while improving reconstruction accuracy.
3Quantity of substance
If moderate bit depths are used with nonlinear transfer functions, then data compression is improved, but quantization precision deteriorates in certain luminance regions
Solution Approach 1:
The patent addresses quantization precision issues by applying parameter changes to the chroma values. Specifically, it adjusts the chroma values based on the local luminance characteristics and the nonlinear transfer function properties. By modifying the chroma parameters in regions where quantization errors are most problematic (steep slope regions), the system maintains acceptable color accuracy even with moderate bit depths, thus preserving data compression benefits while mitigating quantization precision loss.
Data Source
AI summary
One embodiment of the present invention sets forth a technique for correcting color values. The technique includes downsampling first color space values to generate downsampled color space values, upsampling the downsampled color space values to generate second color space values, and determining a first new value for at least one component value included in the downsampled color space values based on a first component value included in the first color space values, a second component value included in the second color space values, and an approximation of a nonlinear transfer function. The technique further includes determining that a first color component value associated with the first new value is outside of a color space range, and determining a second new value for the at least one component value, where the first color component associated with the second new value is within the color space range.