Perceptual Weighting Matrices for Video Transform Coefficients
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding systems face challenges in achieving high compression efficiency while maintaining video quality, as quantization often results in artifacts at low bit rates, affecting user experience.
Innovation Solution
The implementation of perceptually designed weighting parameters for transform coefficients, optimized for the human visual system, which are applied before or after quantization to reduce bit rate without compromising video quality, by selecting and refining weighting matrices based on visual importance and coding costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If quantization is used to reduce bit rate, then compression efficiency is improved, but video quality deteriorates due to artifacts
Solution Approach 1:
The patent applies different weighting factors to different transform coefficients based on their perceptual importance. Less important coefficients (those less likely to be perceived by human visual system) are weighted more heavily for reduction, while important coefficients are preserved. This local differentiation allows aggressive compression in imperceptible areas while maintaining quality in critical areas, resolving the contradiction between bit rate reduction and artifact prevention.
Solution Approach 2:
The patent modifies the quantization process by introducing perceptual weighting factors that change the effective quantization step size for different coefficients. By adjusting these weighting parameters based on human visual system characteristics, the system achieves better rate-distortion performance without introducing visible artifacts, thus reducing bit rate while maintaining video quality.
2Quantity of substance
If traditional quantization is applied, then bit rate is reduced, but visual quality deteriorates
Solution Approach 1:
Different regions of the transform coefficient block are treated differently based on perceptual importance. The patent identifies which coefficients contribute most to perceived quality and applies lighter weighting to those regions while applying heavier weighting to less important regions. This localized quality preservation allows significant bit rate reduction without degrading visual quality in critical areas.
Solution Approach 2:
The patent introduces perceptual weighting parameters that modify the quantization characteristics for different coefficients. By changing these parameters based on human visual system properties and content characteristics, the system achieves more efficient bit rate reduction that maintains visual quality, avoiding the quality deterioration associated with traditional uniform quantization.
3Quantity of substance
If weighting parameters are applied to transform coefficients, then bit rate is reduced, but coding complexity increases
Solution Approach 1:
The perceptual weighting matrices are pre-computed and stored based on human visual system characteristics and content properties. Rather than computing complex weightings in real-time during encoding, the system prepares these weighting parameters in advance, significantly reducing the computational complexity during the actual coding process while still achieving bit rate reduction benefits.
Solution Approach 2:
The patent uses pre-defined perceptual weighting matrices that can be selected and applied based on content characteristics. Instead of computing custom weighting parameters for every block, the system copies and applies appropriate pre-computed matrices, reducing computational complexity while maintaining the bit rate reduction effectiveness of perceptual weighting.
Data Source
AI summary
Techniques related to transform coefficient shaping for video encoding are discussed. Such techniques include applying weighting parameters from one or more perceptually-designed matrices of weighting parameters to blocks of transform coefficients to generate weighted transform coefficients and encoding the weighted transform coefficients into a bitstream. The process may be based on sets of perceptually designed matrices of weighting parameters. Classifier outputs may be used to select from the set of perceptually designed matrices a subset of matrices to work with. The latter may be used in a synthesis procedure to develop the final weighting matrix to be used is shaping the transform coefficients.