Video Reshaping Architectures for Efficient HDR Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently integrating image reshaping and coding, particularly with higher bit depths, leading to increased complexity and suboptimal coding efficiency, especially for high-dynamic range (HDR) content.
Innovation Solution
The implementation of normative out-of-loop, in-loop intra-only, and in-loop residual reshaping architectures within video encoders and decoders, along with efficient signaling methods, to optimize compression and reduce complexity by applying reshaping functions in specific coding loops or as metadata.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If traditional video coding is applied to HDR content with higher bit depths, then coding complexity increases, but coding efficiency deteriorates
Solution Approach 1:
The patent segments the video coding process into distinct reshaping and coding stages. The reshaping function transforms HDR content into a format optimized for compression, separating the complex HDR handling from the standard coding pipeline. This segmentation allows each stage to be optimized independently, reducing overall complexity while maintaining efficiency.
Solution Approach 2:
The patent applies reshaping as a preliminary action before standard video coding. By pre-processing HDR content through the reshaping function, the data is transformed into a distribution that is more amenable to conventional coding techniques, thereby improving coding efficiency without requiring complete redesign of the coding pipeline.
2Productivity
If reshaping is integrated into the video coding loop, then coding efficiency improves, but device complexity increases
Solution Approach 1:
The patent merges the reshaping function with existing video coding components, integrating it into the coding loop rather than treating it as a separate preprocessing step. This merging allows the reshaping to work synergistically with prediction, transform, and quantization stages, improving overall coding efficiency while sharing computational resources.
Solution Approach 2:
The reshaping function is designed to be universal and adaptable to different coding modes and content types. It can be applied to various color spaces, bit depths, and coding configurations, making it a multi-functional component that improves efficiency across diverse scenarios without requiring separate implementations for each case.
3Manufacturing precision
If adaptive reshaping is applied to improve compressibility, then image quality improves, but processing complexity increases
Solution Approach 1:
The patent implements adaptive reshaping where the reshaping parameters are dynamically adjusted based on content characteristics such as luminance distribution, color space, and scene complexity. This dynamic adaptation allows the system to optimize image quality for different content types while avoiding unnecessary processing for content that doesn't benefit from reshaping, thereby managing processing complexity.
Solution Approach 2:
The patent changes key parameters of the reshaping function based on input content analysis, including the choice of transfer function (gamma, PQ, HLG), bit depth, and color space. These parameter changes are made adaptively to match the content requirements, improving image quality while the parameter selection process is designed to be computationally efficient.
Data Source
AI summary
Given a sequence of images in a first codeword representation, methods, processes, and systems are presented for integrating reshaping into a next generation video codec for encoding and decoding the images, wherein reshaping allows part of the images to be coded in a second codeword representation which allows more efficient compression than using the first codeword representation. A variety of architectures are discussed, including: an out-of-loop reshaping architecture, an in-loop-for intra pictures only reshaping architecture, an in-loop architecture for prediction residuals, and a hybrid in-loop reshaping architecture. Syntax methods for signaling reshaping parameters, and image-encoding methods optimized with respect to reshaping are also presented.