LFNST Transform Coding for Sub-Partition 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 like VR and AR, has led to a need for more efficient image/video compression techniques to reduce transmission and storage costs.
Innovation Solution
The implementation of an image coding method using Low-Frequency Non-Separable Transform (LFNST) applied to sub-partition blocks, which involves deriving modified transform coefficients and parsing an LFNST index based on specific regions within the current block, optimizing the coding process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional transform coding is used for high-resolution images, then image quality is maintained, but transmission cost and storage cost increase
Solution Approach 1:
The current block is divided into multiple sub-partition blocks, and LFNST is selectively applied to specific sub-partitions based on the presence of transform coefficients. This segmentation approach allows the transform to be applied only where necessary, reducing overall computational complexity while maintaining image quality.
Solution Approach 2:
LFNST is applied selectively to specific regions (sub-partitions) rather than uniformly across the entire block. The transform is applied only to sub-partitions containing transform coefficients, creating locally optimized processing that reduces computational load while preserving important image details.
2Productivity
If LFNST is applied to all blocks, then compression efficiency improves, but computational complexity increases
Solution Approach 1:
LFNST is applied partially rather than universally - specifically to sub-partitions containing transform coefficients. This partial application strategy achieves compression efficiency improvements where needed while avoiding unnecessary computational overhead in regions where the transform would provide minimal benefit.
Solution Approach 2:
The processing is segmented into multiple stages: first identifying sub-partitions with transform coefficients, then applying LFNST only to those specific regions. This segmented approach reduces computational complexity by avoiding redundant transforms in regions where coefficients are absent.
3Measurement precision
If transform coefficients are processed in the entire current block, then coding accuracy improves, but processing time increases
Solution Approach 1:
The method extracts and processes only the relevant portions of the block - specifically sub-partitions containing transform coefficients. By taking out and processing only these necessary regions rather than the entire block, the method maintains coding accuracy for important features while significantly reducing processing time.
Solution Approach 2:
Instead of processing the entire current block, the method applies LFNST partially to only those sub-partitions that contain transform coefficients. This partial processing approach maintains sufficient coding accuracy for regions that need it while eliminating wasteful processing time in regions where no coefficients exist.
Data Source
AI summary
An image decoding method includes a step of deriving a modified transform coefficient, wherein the step of deriving the modified transform coefficient includes the steps of: deriving a first variable indicating whether the transform coefficient exists in an area excluding a DC location of the current block; parsing an LFNST index based on the derivation result; and deriving the modified transform coefficient based on the LFNST index and an LFNST matrix, and based on the fact that the current block is divided into a plurality of sub-partition blocks, the LFNST index can be parsed without deriving the first variable.