Image Transform Coding with LFNST Index Signaling

Resolve Bottlenecks,
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 highly efficient image/video compression technique to reduce transmission and storage costs.

Innovation Solution

An image coding method that includes deriving a transform coefficient for a current block based on residual information, signaling an LFNST index for the luma component, and applying a modified transform coefficient using an LFNST matrix, with optional signaling of an MTS index for dual tree chroma or ISP mode blocks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional image coding methods are used for high-resolution images, then image quality is maintained, but transmission cost and storage cost increase

Engineering Contradiction:

Improveimage qualityVSAvoidbit amount

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The image block is divided into multiple sub-blocks, and different transform kernels are applied to different sub-blocks. This segmentation allows for more precise representation of local image characteristics, improving compression efficiency while maintaining image quality.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different transform kernels are selected for different sub-blocks based on their local characteristics. This local adaptation optimizes the transform representation for each region, achieving better compression ratios without sacrificing overall image quality.

Inventive Principle:

Principle #3Local quality

2Measurement precision

If transform coefficients are fully signaled for all blocks, then coding precision is improved, but signaling overhead increases

Engineering Contradiction:

Improvecoding precisionVSAvoidsignaling overhead

Core Design Contradiction:

Measurement precisionVSLoss of information

Solution Approach 1:

The LFNST index is extracted and signaled separately from the main transform coefficient data. This allows for efficient representation of the low-frequency non-separable transform application, reducing overall signaling overhead while maintaining coding precision where needed.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The LFNST is applied selectively only to certain sub-blocks based on their characteristics, rather than uniformly to all blocks. This partial application reduces the number of transform indices that need to be signaled, lowering overhead while maintaining precision for blocks that benefit most.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS20260106986A1Transform-based image coding method and apparatus therefor

Publication Date: 2026.04.16 LG ELECTRONICS INC

AI summary

An image decoding method according to the present document comprises the steps of: deriving a transform coefficient for a current block on the basis of residual information; when LFNST is applied to the current block, receiving an LFNST index related to an LFNST matrix; and deriving a modified transform coefficient for the current block on the basis of the LFNST matrix, wherein the LFNST index may be signaled for a luma component of the current block.