Sub-Block Intra Prediction for High-Efficiency Video Encoding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The increasing demand for high-resolution and high-quality images, particularly stereo-scopic content, leads to higher data volumes, increasing transmission and storage costs, and existing image compression technologies are inadequate for efficient encoding and decoding of such data.

Innovation Solution

A method and device for performing intra prediction in sub-block units, utilizing a weighted sum operation between reference modes, and selectively determining inverse transforms based on block distances, to improve encoding efficiency and reduce data storage requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If intra prediction is performed in sub-block units with multiple reference modes, then prediction accuracy and encoding efficiency are improved, but complexity of deriving reference modes and determining inverse transforms increases

Engineering Contradiction:

Improveencoding efficiencyVSAvoidcomplexity of deriving reference modes

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The current block is divided into multiple sub-blocks, each undergoing independent intra prediction with its own reference modes. This segmentation allows finer-grained prediction control and improved accuracy while managing complexity through structured processing of smaller units

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different sub-blocks within the same block can have different reference modes selected based on their specific characteristics and neighboring block relationships. This local quality approach optimizes prediction for each sub-block individually rather than applying a uniform approach throughout the entire block

Inventive Principle:

Principle #3Local quality

2Reliability

If all intra prediction modes for all sub-blocks are stored, then complete prediction information is available, but storage bandwidth and data volume increase

Engineering Contradiction:

Improveprediction information completenessVSAvoidstorage bandwidth

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

Only the intra prediction mode of the representative sub-block is extracted and stored, while the prediction modes of other sub-blocks are derived through inheritance or prediction relationships. This extraction approach maintains essential prediction information while significantly reducing the quantity of data that needs to be stored and transmitted

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The representative block's intra prediction mode serves multiple sub-blocks through inheritance relationships, allowing a single stored value to provide prediction information for multiple blocks. This multi-functionality reduces the total storage requirement while maintaining prediction capability across all sub-blocks

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS12542911B2Video signal encoding/decoding method based on intra prediction in sub-block units, and recording medium for storing bit-stream

Publication Date: 2026.02.03 KT CORP

AI summary

A video encoding/decoding method according to the present disclosure may comprise the steps of: inducing a reference mode of respective neighboring blocks within a current block; and performing, on the basis of at least one of reference modes, intra prediction on a sub-block, in the current block, that is smaller than the current block.