LFNST Chroma Transform Selection for Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images/videos, such as 4K and 8K ultra high definition, leads to higher bit rates, resulting in increased transmission and storage costs. Additionally, the need for efficient compression techniques is exacerbated by the rise of immersive media like virtual and augmented reality.
Innovation Solution
The proposed solution involves an image coding method and apparatus that utilizes the Low-Frequency Non-Separable Transform (LFNST) to enhance coding efficiency. This method includes deriving an LFNST transform set by borrowing an intra mode of a luma block in Cross-Component Linear Model (CCLM) mode, and selecting an LFNST matrix based on the derived set and index to generate transform coefficients for chroma blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If high-resolution and high-quality images/videos are transmitted or stored, then image quality is improved, but transmission cost and storage cost increase
Solution Approach 1:
The patent applies parameter changes by adapting transform matrices based on prediction modes and block types. Different transform matrices are selected from predefined sets depending on the intra prediction mode (e.g., angular, planar, DC) and block characteristics, allowing optimal compression parameters to be adjusted for each region rather than using a fixed transformation approach
Solution Approach 2:
The system dynamically selects transform matrices based on the specific prediction mode and block type encountered during encoding. The transform matrix selection is not static but adapts in real-time based on the content characteristics, enabling the system to optimize compression efficiency for different image regions and prediction scenarios
2Productivity
If conventional compression techniques are used, then transmission and storage costs are reduced, but coding efficiency for high-resolution content decreases
Solution Approach 1:
The patent segments the transform matrix selection process into multiple predefined sets, each associated with specific prediction modes or block types. This segmentation allows the complex task of optimizing transforms for all possible scenarios to be broken down into manageable categories, where appropriate matrices are pre-selected based on block characteristics
Solution Approach 2:
The system manages complexity by pre-defining multiple transform matrix sets and using simple selection criteria based on prediction modes. Rather than computing optimal transforms in real-time, the system changes parameters by selecting from pre-computed matrices, reducing computational complexity while maintaining high compression efficiency
3Productivity
If transform matrices are selected based on prediction modes, then coding efficiency is improved, but processing complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-defining multiple transform matrix sets before the actual encoding process. These matrices are prepared in advance and organized into sets that can be quickly selected based on prediction modes, eliminating the need for complex real-time computation while maintaining optimization
Solution Approach 2:
The system simplifies the transform selection process by changing parameters through direct indexing into pre-defined matrix sets based on prediction mode identifiers. This approach replaces complex optimization algorithms with simple parameter lookup and selection, reducing processing complexity while preserving coding efficiency
Data Source
AI summary
An image decoding method according to the present document can comprise the steps of: acquiring, from a bitstream, intra prediction mode information and an LFNST index; deriving, as a cross-component linear model (CCLM) mode, an intra prediction mode of a chroma block on the basis of the intra prediction mode information; changing the intra prediction mode of the chroma block from the CCLM mode to an intra prediction mode of a luma block corresponding to the chroma block; determining an LFNST set, including LFNST matrices, on the basis of the intra prediction mode of the luma block; selecting one of the LFNST matrices on the basis of the LFNST set and the LFNST index; and deriving transform coefficients for the chroma block on the basis of the selected LFNST matrix.