Adaptive Quantization Matrix for Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face challenges in improving coding efficiency, enhancing image quality, and reducing circuit scale and processing amount, particularly in adapting to varying block sizes during encoding and decoding processes.
Innovation Solution
An encoder and decoder that generate and utilize specific quantization matrices based on the size of the current block, allowing for adaptive quantization and inverse quantization to improve image quality and processing efficiency by adjusting matrix elements in the low frequency domain.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single quantization matrix is used for all block sizes, then device complexity is reduced, but coding efficiency and image quality deteriorate
Solution Approach 1:
The patent segments the quantization process by creating separate quantization matrices for different block sizes (e.g., 4×4, 8×8, 16×16, 32×32 blocks). This allows each block size to have its own optimized quantization characteristics, improving coding efficiency without requiring complex adaptive matrix generation for every block.
Solution Approach 2:
The patent applies local quality by using different quantization matrices for different frequency domains within blocks. Specifically, it uses a first quantization matrix for low-frequency coefficients and a second quantization matrix for high-frequency coefficients, allowing optimized quantization for each frequency region while maintaining manageable device complexity.
2Manufacturing precision
If adaptive quantization matrices are generated for each block size, then coding efficiency improves, but processing amount increases
Solution Approach 1:
The patent prepares multiple quantization matrices in advance for different block sizes and frequency domains before actual encoding begins. These pre-computed matrices are stored and directly applied during encoding, eliminating the need for real-time matrix generation and reducing processing amount during the actual encoding operation.
Solution Approach 2:
The patent changes the quantization parameters by selecting different pre-prepared quantization matrices based on the current block size and frequency domain characteristics. This allows adaptive optimization for each block without requiring complex real-time calculations, thus improving coding efficiency while keeping processing amount manageable through parameter selection rather than computation.
3Manufacturing precision
If different quantization matrices are used for low and high frequency domains, then image quality improves, but device complexity increases
Solution Approach 1:
The patent segments the frequency domain into low-frequency and high-frequency regions, applying different quantization matrices to each segment. This segmentation allows optimized quantization for each frequency region, improving image quality by preserving important low-frequency information while adequately handling high-frequency details, without requiring a completely new complex quantization architecture.
Data Source
AI summary
An encoder includes circuitry and memory coupled to the circuitry. The circuitry, in operation: generates (i) a first quantization matrix for transform coefficients included in a current block to be processed and (ii) a second quantization matrix for transform coefficients included in a low frequency domain among the transform coefficients included in the current block; and quantizes the transform coefficients included in the current block using at least one of the first quantization matrix or the second quantization matrix, in accordance with a size of the current block.