CCSO Filtering with Asymmetric Quantization for Video Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in effectively reducing redundancy in uncompressed digital video signals, particularly in cross-component sample offset filtering, which affects storage and transmission efficiency.
Innovation Solution
Implementing an asymmetric quantizer for cross-component sample offset filtering, where a quantizer is selected based on syntax elements in the video bitstream to apply CCSO filters, allowing for efficient quantization of sample deltas and improved filtering of reconstructed samples.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a symmetric quantizer is used for CCSO filtering, then the quantization process is simple and standard, but the filtering effectiveness and video quality are insufficient
Solution Approach 1:
The patent applies asymmetry by introducing an asymmetric quantizer that uses different quantization step sizes for positive and negative sample deltas. Specifically, the quantizer uses a first quantization step size for positive deltas and a second quantization step size for negative deltas, where these step sizes are derived from different scaling factors. This asymmetric approach allows the quantizer to better capture the statistical characteristics of video residuals, improving filtering effectiveness while maintaining manageable complexity through structured parameter derivation.
2Measurement precision
If asymmetric quantization is applied to improve filtering quality, then video reconstruction quality improves, but the complexity of quantizer configuration and signaling increases
Solution Approach 1:
The patent implements parameter changes by deriving quantization step sizes from scaling factors that are themselves derived from standard video coding parameters. The first and second quantization step sizes are calculated using scaling factors that can be obtained from existing quantization parameters and color component relationships. This approach allows the asymmetric quantizer to adapt to different video content and coding conditions without requiring entirely new configuration mechanisms, thus improving video quality while controlling configuration complexity.
Solution Approach 2:
The asymmetric quantizer is designed to work within the existing video coding framework, serving multiple functions: it handles different color components (luma and chroma), adapts to various quantization scenarios, and integrates with existing syntax elements. The quantizer can be configured using existing coding parameters and can operate in different modes depending on the video content and coding conditions, making it a universal solution that improves quality without requiring completely separate configuration systems.
3Adaptability or versatility
If multiple quantizers are supported with flexible signaling, then adaptability to different video content improves, but the bitstream complexity and signaling overhead increase
Solution Approach 1:
The patent implements dynamics by allowing the quantizer configuration to adapt based on the specific coding conditions and video content. The asymmetric quantizer can dynamically adjust its behavior by selecting different scaling factors and quantization step sizes based on the color component, quantization parameter, and other context information. This dynamic adaptation enables the system to optimize filtering performance for different video scenarios while using a unified quantizer structure, avoiding the need for multiple completely different quantizer types and reducing signaling complexity.
Data Source
AI summary
This disclosure describes aspects of advanced video coding technologies and is particular related to cross-component sample offset (CCSO) filtering of reconstructed samples. For example, a quantizer for quantizing sample deltas in order to apply a CCSO filter may be constructed as an asymmetric quantizer with respect to zero sample delta. Either a symmetric or an asymmetric quantizer may be selected for processing sample deltas prior to applying a CCSO filter. An asymmetric quantizer may relate to a corresponding symmetric quantizer by a non-zero offset. Signaling of a selection of an asymmetric quantizer may be provided separately from or dependent on a signaling of a symmetric quantizer.