Adaptive Residual Quantization for Backward-Compatible Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding technologies struggle to balance data size reduction with visual quality and processing efficiency, particularly in hybrid backward-compatible coding systems.
Innovation Solution
A method involving downsampling an input video, encoding it with a base codec, generating residuals, and applying adaptive quantization to these residuals based on their characteristics, allowing for varying step-widths and quantization parameters to optimize compression efficiency and visual quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a fixed quantization matrix is used for all frequency components, then the encoding process is simple and fast, but visual quality deteriorates because perceptually important components are not allocated sufficient data
Solution Approach 1:
The quantization matrix is made dynamic and adaptive rather than fixed. The encoder adjusts quantization parameters based on the actual frequency content and statistical characteristics of each video block, allowing the quantization strength to vary dynamically to optimize both compression efficiency and visual quality.
Solution Approach 2:
The quantization parameters are changed adaptively based on the statistical properties of the transform coefficients. By analyzing the distribution characteristics of different frequency components and adjusting quantization steps accordingly, the system allocates more bits to perceptually important components while maintaining encoding efficiency.
2Quantity of substance
If quantization compression is applied to reduce data size, then storage and transmission requirements are reduced, but processing complexity increases due to adaptive quantization operations
Solution Approach 1:
Statistical analysis and quantization parameter determination are performed in advance during the encoding process. By pre-calculating appropriate quantization parameters based on coefficient statistics before actual quantization, the system reduces online processing complexity while achieving optimal compression ratios.
Solution Approach 2:
The quantization process utilizes the inherent statistical properties of the video data itself to determine optimal compression parameters. The system performs self-adaptive quantization by analyzing its own input data characteristics, eliminating the need for complex external control mechanisms.
3Productivity
If uniform quantization is applied to all residual elements, then the encoding process is computationally efficient, but compression efficiency deteriorates because different frequency components have different impacts on visual quality
Solution Approach 1:
Different quantization strengths are applied to different frequency components based on their local importance to visual quality. Low-frequency components that have greater impact on perceived quality receive finer quantization (smaller quantization steps), while high-frequency components receive coarser quantization, optimizing the overall compression efficiency.
Solution Approach 2:
Quantization parameters are varied according to the statistical characteristics of different frequency components. By changing quantization steps based on the distribution and magnitude of transform coefficients in different frequency bands, the system achieves better compression efficiency while maintaining encoding practicality.
Data Source
AI summary
According to aspects of the invention there is provided a method of encoding an input video into a plurality of encoded streams, wherein the encoded streams may be combined to reconstruct the input video. There may be provided an encoding method comprising: receiving an input video; downsampling the input video to create a downsampled video; instructing an encoding of the downsampled video using a base encoder to create a base encoded stream; instructing a decoding of the base encoded stream using a base decoder to generate a reconstructed video; comparing the reconstructed video to the downsampled video to create a first set of residuals; and, encoding the first set of residuals to create a first level encoded stream, including: applying a transform to the first set of residuals to create a first set of coefficients; applying a quantization operation to the first set of coefficients to create a first set of quantized coefficients; and applying an encoding operation to the first set of quantized coefficients, wherein applying the quantization operation comprises: adapting the quantization based on the first set of coefficients to be quantized, including varying a step-width used for different ones of the first set of coefficients, wherein a first set of parameters derived from the adapting is signalled to a decoder to enable dequantization of the first set of quantized coefficients.