Adaptive Constant-Luminance Video Coding for HDR WCG Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding techniques, such as those in HEVC, are not fully optimized for High Dynamic Range (HDR) and Wide Color Gamut (WCG) video data, leading to inefficiencies in compression due to the lack of flexibility in existing color transforms like Constant Luminance (CL), which do not adequately address the spatio-temporal dynamics of HDR/WCG content.
Innovation Solution
The introduction of Adaptive Constant-Luminance (ACL) techniques with four scaling factors for chroma components, allowing for more flexible and accurate representation of HDR/WCG data, which can be combined with other techniques like transfer functions and quantization, without significant complexity increase.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If Constant Luminance (CL) transform is used for HDR/WCG video coding, then the coding process is simple, but the compression efficiency is insufficient due to lack of flexibility
Solution Approach 1:
The patent introduces adaptive scaling factors that dynamically adjust the chroma components based on local picture characteristics. Instead of using fixed scaling factors in traditional CL transform, the encoder selects from multiple predefined scaling factor sets (e.g., sfCb_0 to sfCb_3, sfCr_0 to sfCr_3) based on chroma variance and saturation metrics, making the transform adaptable to different content regions while maintaining reasonable computational complexity.
Solution Approach 2:
The patent changes the parameters of the color transform by introducing multiple scaling factor options for chroma components. The encoder calculates chroma variance and saturation to determine which scaling factor set to apply, thereby changing the transform parameters adaptively rather than using fixed parameters. This allows the system to optimize compression efficiency for different picture regions without significantly increasing complexity.
2Measurement precision
If adaptive scaling factors are introduced to improve HDR/WCG representation accuracy, then coding precision increases, but computational complexity increases
Solution Approach 1:
The patent introduces adaptive scaling factors that modify the chroma components based on local picture characteristics. By calculating chroma variance and saturation metrics, the system selects from predefined scaling factor sets to achieve higher coding precision for HDR/WCG content while keeping the computational overhead manageable through constrained selection from a limited set of scaling factors.
Solution Approach 2:
The system dynamically adjusts chroma scaling factors based on local picture characteristics such as chroma variance and saturation. This dynamic adaptation allows the system to improve precision where needed (in regions with high chroma variation) while avoiding unnecessary computations in regions where fixed scaling would suffice, thereby balancing precision and complexity.
3Productivity
If traditional color transforms are used for video coding, then the processing is efficient, but they do not adequately address spatio-temporal dynamics of HDR/WCG content
Solution Approach 1:
The patent makes the color transform dynamic by introducing adaptive scaling factor selection based on chroma variance and saturation calculations. The system adjusts scaling factors for different picture regions and can switch between different scaling factor sets to match the spatio-temporal characteristics of HDR/WCG content, thereby improving adaptability while maintaining processing efficiency through constrained adaptation options.
Solution Approach 2:
The patent changes the transform parameters adaptively by selecting from multiple scaling factor sets based on local picture characteristics. This allows the system to adapt to different HDR/WCG content characteristics (such as high saturation regions or areas with significant chroma variation) while maintaining processing efficiency through a limited set of predefined scaling options rather than fully adaptive parameter optimization.
Data Source
AI summary
A device may determine, based on data in a bitstream, a luma sample (Y) of a pixel, a Cb sample of the pixel, and the Cr sample of the pixel. Furthermore, the device may obtain, from the bitstream, a first scaling factor and a second scaling factor. Additionally, the device may determine, based on the first scaling factor, the Cb sample for the pixel, and Y, a converted B sample (B′) for the pixel. The device may determine, based on the second scaling factor, the Cr sample for the pixel, and Y, a converted R sample (R′) for the pixel. The device may apply an electro-optical transfer function (EOTF) to convert Y′, R′, and B′ to a luminance sample for the pixel, a R sample for the pixel, and a B sample for the pixel, respectively.