Adaptive HDR Video Reshaping for HEVC Compatibility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding and decoding schemes, such as MPEG-4 AVC and HEVC, are not designed to handle High Dynamic Range (HDR) and Wide Color Gamut (WCG) video directly, leading to challenges in quantizing 16-bit floating point color values into 10-bit integer values, resulting in distortion and loss of detail.
Innovation Solution
A system and method that regenerates HDR video data by extracting a self-referential metadata structure specifying a video data reshaping transfer function, applying it to decoded video data to redistribute color and intensity information based on the content's characteristics, and encoding it using non-HDR/ non-WCG encoders like H.265 HEVC, with inverse operations performed at the decoder to restore HDR/WCG data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If uniform quantization is applied to convert 16-bit floating point HDR color values to 10-bit integer values, then the encoding process becomes simpler, but significant distortion and loss of detail occur
Solution Approach 1:
The patent applies a coding transfer function that changes the parameter space from linear 16-bit floating point HDR values to non-linear 10-bit integer values. This transfer function compresses the dynamic range by mapping the broad HDR range (0 to 10,000 nits) to a narrower SDR range (0 to 255 or 0 to 1023), enabling uniform quantization while preserving perceptual detail through non-uniform distribution of quantization levels across different luminance ranges.
Solution Approach 2:
The patent employs a dynamic coding transfer function that adapts to the content characteristics of the input video. Instead of using a fixed static mapping, the system analyzes the input HDR video data to determine appropriate transfer function parameters, allowing the quantization to dynamically adjust to the specific luminance and color distribution in each frame, thereby optimizing the balance between compression and detail preservation.
2Ease of manufacture
If a fixed coding transfer function is used to map HDR values to SDR values, then the encoding process is simpler, but poor allocation of quantization levels results in contouring and banding
Solution Approach 1:
The patent transitions from a static fixed transfer function to a dynamic adaptive transfer function. The system analyzes the input HDR video data characteristics (such as luminance distribution, color gamut, and scene content) to automatically select or compute appropriate transfer function parameters for each frame or sequence. This dynamic adaptation ensures that quantization levels are optimally allocated according to the actual content, preventing contouring and banding artifacts while maintaining encoding simplicity.
Solution Approach 2:
The patent incorporates feedback mechanisms where the encoder analyzes the input HDR video data to determine the characteristics of the scene (luminance range, color distribution, etc.) and uses this information to adjust the coding transfer function parameters. This feedback loop allows the system to optimize the quantization allocation in real-time based on the actual content, ensuring high quantization quality without requiring complex manual configuration.
3Adaptability or versatility
If HDR video is encoded using existing SDR video compression algorithms, then compatibility is improved, but the broad dynamic range cannot be adequately represented
Solution Approach 1:
The patent introduces a coding transfer function as an intermediary between the HDR input space and the SDR compression algorithm output space. This transfer function acts as a mediator that converts the broad dynamic range HDR values into a compressed non-linear SDR representation that is compatible with existing compression algorithms. The transfer function preserves the essential dynamic range information by applying non-uniform quantization that maintains perceptual accuracy, while the compatibility with SDR algorithms is maintained through standard quantization and encoding procedures.
Solution Approach 2:
The patent changes the parameter representation from HDR (16-bit floating point, broad dynamic range) to SDR (10-bit integer, narrower dynamic range) through a carefully designed coding transfer function. This parameter transformation enables HDR video to be processed by existing SDR compression algorithms while preserving the essential dynamic range information through non-uniform quantization that allocates more precision to critical luminance ranges and compresses less critical ranges appropriately.
Data Source
AI summary
A system and method for regenerating high dynamic range (HDR) video data from encoded video data, extracts, from the encoded video data, a self-referential metadata structure specifying a video data reshaping transfer function. The video data reshaping transfer function is regenerated using data from the metadata structure and the extracted reshaping transfer function is used to generate the HDR video data by applying decoded video data values to the reshaping transfer function.