Chroma Intra Prediction Scaling Factor for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing LM mode chroma intra prediction method in video coding is computationally intensive and requires complex calculations, including large multipliers and buffer storage, which complicates the derivation of prediction samples.
Innovation Solution
A simplified method for chroma intra prediction based on a linear model using derived co-located luma pixels scaled by a scaling factor, where the scaling factor incorporates a division factor related to the internal bit depth and variance-like values, reducing the complexity of operations such as a1s*lmDiv by modifying the shift operations and look-up table entries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the existing LM mode chroma intra prediction method is used, then prediction accuracy is improved, but computational complexity increases due to large multipliers and buffer storage requirements
Solution Approach 1:
The patent changes the parameters of the linear model by introducing a scaling factor that incorporates a division factor and a scaled covariance-like value. This transformation modifies the computational requirements by replacing large multipliers with shifted operations and look-up table entries, thereby reducing computational complexity while maintaining prediction accuracy
Solution Approach 2:
The patent substitutes complex arithmetic operations (multiplication and division) with simpler operations (bit shifting and table lookup). The scaled covariance-like value and division factor are pre-computed and stored in look-up tables, replacing the need for real-time complex calculations with simpler mechanical operations
2Measurement precision
If complex calculations including large multipliers are performed, then prediction precision is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary computation by pre-calculating and storing scaled covariance-like values and division factors in look-up tables during system initialization or buffer allocation. This preliminary action eliminates the need for complex real-time calculations during video decoding, significantly reducing processing time while maintaining prediction precision
Solution Approach 2:
The patent replaces time-consuming arithmetic operations (multiplication and division) with faster bit shifting operations and table lookup operations. The scaled covariance-like value is computed using right shift operations, and the division factor is obtained through table lookup, both of which are computationally much lighter than traditional arithmetic operations
3Measurement precision
If high precision arithmetic is used, then calculation accuracy is improved, but device complexity increases
Solution Approach 1:
The patent changes the parameter representation by introducing a scaling factor that transforms the calculation into a form suitable for fixed-point arithmetic. The scaled covariance-like value and division factor are designed to work with standard integer operations, eliminating the need for high precision floating-point arithmetic while maintaining sufficient calculation accuracy
Solution Approach 2:
The patent uses look-up tables with pre-computed values as disposable computational resources. These tables are created once during initialization and then repeatedly queried during processing, replacing the need for complex arithmetic operations. The table entries serve as ready-to-use computational primitives that eliminate the need for high precision arithmetic hardware
Data Source
Figure 1A~1B
Figure 2
AI summary
A method and apparatus for chroma intra prediction based on reconstructed luma pixels and chroma pixels are disclosed. The chroma intra prediction is based on a linear model of derived co-located current luma pixels of the current luma block scaled by a scaling factor. The scaling factor comprises a product term of a division factor and a scaled covariance-like value associated with neighboring reconstructed luma and chroma pixels of a current block. The division factor is related to a first data range divided with rounding by a scaled variance-like value associated with the neighboring reconstructed luma pixels of the current block. The scaled covariance-like value, the first data range, or both of the scaled covariance-like value and the first data range are dependent on the internal bit depth, with which the chroma signal is processed during video coding process, according to an embodiment of the present invention.