Chroma-From-Luma Prediction Using Derived Scaling Factors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression techniques do not effectively utilize the correlation between luma and chroma components to reduce data requirements, leading to inefficient encoding and decoding processes.
Innovation Solution
A method for predicting chroma values from luma values using a derived scaling factor, which involves reconstructing luma blocks, determining average luminance and chrominance values, applying a scaling factor to difference values, and combining these to form a chroma-from-luma prediction block.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If chroma blocks are encoded independently without utilizing luma-chroma correlation, then the encoding process is simple, but the data compression efficiency is poor
Solution Approach 1:
The patent merges the encoding of chroma blocks with luma block information by predicting chroma values from luma values using a linear model. This combines previously separate encoding processes into a unified approach that leverages the correlation between luma and chroma components, improving compression efficiency without requiring completely independent encoding of both components.
Solution Approach 2:
The patent introduces a scaling factor parameter that transforms the relationship between luma and chroma components. By adjusting this scaling factor (either derived from neighboring blocks or explicitly signaled), the system optimizes the prediction of chroma values from luma values, enabling better compression efficiency through parameter optimization rather than fundamental process redesign.
2Measurement precision
If a scaling factor is explicitly signaled in the bitstream, then the prediction accuracy is improved, but the bitstream complexity and data requirements increase
Solution Approach 1:
The system performs self-service by deriving the scaling factor from neighboring block information that is already available during the decoding process. Instead of requiring external signaling of the scaling factor, the decoder computes it autonomously from surrounding chroma and luma block data, achieving accurate prediction without additional bitstream overhead.
Solution Approach 2:
The patent uses neighboring block data as an intermediary to indirectly determine the scaling factor. Rather than directly signaling the scaling factor itself, the system uses information from adjacent blocks (both chroma and luma) as a mediator to compute the scaling factor, reducing the need for direct signaling while maintaining prediction accuracy.
3Productivity
If chroma values are predicted from luma values using a derived scaling factor, then the coding efficiency is improved, but the computational complexity increases
Solution Approach 1:
The patent applies partial action by using a simplified linear model for prediction rather than complex neural networks or advanced algorithms. The prediction formula uses basic arithmetic operations (multiplication by scaling factor, addition of offset) on luma values to generate chroma predictions, achieving good coding efficiency without the computational burden of more complex prediction models.
Solution Approach 2:
The system manages computational complexity by changing parameters such as using a fixed or derived scaling factor rather than complex adaptive models. The linear prediction model with adjustable scaling factor and offset parameters provides a balance between computational simplicity and prediction effectiveness, avoiding the need for computationally intensive while maintaining coding efficiency.
Data Source
AI summary
Chroma-from-luma (CfL) intra prediction mode is described that allows for a derived scaling factor. A luma block is reconstructed from an encoded bitstream. An average luminance value for luma pixel values and difference values between the luma pixel values and the average luminance value are determined. An average chrominance value for a chroma block is determined. From a flag in the encoded bitstream, whether a scaling factor for the mode is explicitly signaled or should be derived is determined. Deriving the scaling factor uses pixel values of at least one neighboring block, and otherwise the scaling factor is determined from the encoded bitstream. The scaling factor is applied to the difference values to obtain scaled difference values, a CfL prediction block is obtained by adding the average chrominance value to the scaled difference values, and the chroma block is reconstructed by adding the CfL prediction block to a residual block.