Image Coding with Color-Specific LFNST for High-Resolution Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images/videos, including immersive media, necessitates a highly efficient image/video compression technique to reduce transmission and storage costs.
Innovation Solution
An image coding method and apparatus that utilizes LFNST (Lifting Factorized Nested Transform) based on a tree type, with flag variables for each color component, to enhance compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution and high-quality image/video data is transmitted or stored using conventional methods, then image quality and resolution are maintained, but transmission cost and storage cost increase
Solution Approach 1:
The patent extracts and transmits only the essential visual information by applying transforms (DCT, DST, LFNST) to convert image data into transform coefficients, where only significant coefficients are retained and transmitted. This extraction of essential information maintains image quality while reducing the amount of data that needs to be transmitted or stored, thereby lowering transmission and storage costs.
2Loss of energy
If conventional compression techniques are applied to high-resolution images/videos, then transmission cost is reduced, but compression efficiency is insufficient for ultra high definition content
Solution Approach 1:
The patent implements dynamic transform selection where the transform type (DCT, DST, or LFNST) is adaptively chosen based on the characteristics of the current block, such as prediction mode and block content. This dynamic adaptation allows the compression system to optimize for each specific block, achieving higher compression efficiency for ultra HD content while maintaining quality and reducing transmission costs.
Solution Approach 2:
The patent changes the transform parameters by selecting different transform types (DCT-2, DST-7, DST-10, LFNST) based on the specific characteristics of each image block. By varying the transform parameter (transform type) according to block characteristics, the system achieves superior compression efficiency compared to using a single fixed transform, thereby improving productivity while reducing transmission costs.
3Productivity
If LFNST is applied to all color components uniformly, then compression efficiency may improve, but adaptability to different tree types and color formats is reduced
Solution Approach 1:
The patent applies LFNST selectively to specific color components (luma or chroma) based on the tree type of the current block. For single-tree structures, LFNST is applied only to luma components, while for dual-tree chroma structures, it is applied to chroma components. This localized application of LFNST according to the specific characteristics of each block maintains adaptability to different tree types while achieving compression efficiency improvements where appropriate.
Data Source
AI summary
An image decoding method according to the present document may comprise the steps of: receiving image information including residual information and an LFNST index from a bitstream; deriving a transform coefficient on the basis of the residual information; deriving a flag variable related to whether LFNST is applied to a current block on the basis of the LFNST index; and performing the LFNST on the basis of the flag variable and the transform coefficient, wherein the flag variable is derived for each color component of the current block.