Sub-Partition LFNST Coding for High-Resolution Image Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images/videos, particularly in immersive media formats, necessitates a more efficient image/video compression technique to reduce transmission and storage costs.
Innovation Solution
An image coding method and apparatus utilizing LFNST (Large Frequency Non-separable Transform) applied to sub-partition transform blocks, with shared intra prediction modes and LFNST matrices across sub-partitions, enhancing transform index coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution and high-quality images/videos are transmitted or stored, then image quality and resolution are improved, but transmission cost and storage cost increase
Solution Approach 1:
The patent divides the current block into multiple sub-partition transform blocks (e.g., four quadrants) and applies independent LFNST transforms to each sub-partition. This segmentation allows finer-grained frequency transformation, improving compression efficiency while maintaining image quality, thereby reducing the bit amount needed for transmission and storage.
Solution Approach 2:
The patent applies different LFNST transform matrices selectively to different sub-partitions based on their specific characteristics and requirements. Each sub-partition can receive the transform that best suits its local content, optimizing the balance between compression efficiency and quality preservation for each region.
2Productivity
If conventional transform coding is used, then coding process is simple, but compression efficiency is insufficient for high-resolution images
Solution Approach 1:
The patent segments the transform block into sub-partitions and applies LFNST to each, increasing compression efficiency through finer frequency analysis. The complexity is managed by processing smaller sub-blocks independently rather than applying a single complex transform to the entire block.
Solution Approach 2:
The patent changes the transform parameters by selecting from multiple available LFNST matrices based on the intra-prediction mode and sub-partition characteristics. This adaptive parameter selection optimizes compression efficiency while keeping the actual transform computation relatively simple through a finite set of predefined matrices.
3Productivity
If LFNST is applied to the entire current block, then transform efficiency is improved, but flexibility and adaptability to different sub-partition characteristics are reduced
Solution Approach 1:
The patent segments the current block into sub-partitions and applies LFNST to each individually, maintaining transform efficiency while enabling adaptability to different sub-partition characteristics through independent processing of each segment.
Solution Approach 2:
The patent applies different LFNST transform matrices selectively to different sub-partitions based on their specific characteristics and requirements. Each sub-partition can receive the transform that best suits its local content, optimizing the balance between compression efficiency and quality preservation for each region.
Data Source
AI summary
An image decoding method according to the present document may comprise the steps of: if a current block is partitioned into sub-partition transform blocks, deriving a prediction sample of the current block on the basis of intra prediction mode information; determining an LFNST set including LFNST matrices on the basis of an intra prediction mode derived from the intra prediction mode information; selecting one of the LFNST matrices on the basis of the LFNST set and the LFNST index; deriving transform coefficients for the sub-partition transform blocks on the basis of the selected LFNST matrix; and deriving residual samples for the current block on the basis of the transform coefficients.