MIP Intra Prediction Using Partial Reference Samples

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The increasing demand for high-resolution and high-quality images leads to a surge in transmission and storage costs due to higher bit rates, necessitating a high-efficiency image compression technology.

Innovation Solution

An image encoding/decoding method utilizing matrix-based intra prediction (MIP) that generates prediction blocks using only a part of reference samples, such as the left or top reference samples, to enhance encoding/decoding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If high-resolution and high-quality images are transmitted, then image quality is improved, but transmission cost increases

Engineering Contradiction:
Improveimage qualityVSAvoidtransmission cost
Core Design Contradiction:
Measurement precisionVSLoss of energy

Solution Approach 1:

The patent extracts and transmits only the residual information (difference between original and predicted blocks) rather than the complete high-resolution image data. By separating the prediction component (generated locally at decoder) from the residual component (transmitted), the system maintains image quality while significantly reducing transmission cost.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent creates a predicted copy of the image block at the decoder side using matrix-based intra prediction, which replicates the majority of the image information without transmission. Only the differences (residuals) need to be transmitted, reducing the amount of data while preserving reconstruction quality.

Inventive Principle:
Principle #26Copying

2Measurement precision

If high-resolution and high-quality images are stored, then image quality is improved, but storage cost increases

Engineering Contradiction:
Improveimage qualityVSAvoidstorage cost
Core Design Contradiction:
Measurement precisionVSLoss of energy

Solution Approach 1:

The patent extracts only the essential residual information needed to reconstruct high-quality images, storing the prediction models and parameters locally rather than storing complete high-resolution image data. This separation reduces storage requirements while maintaining image quality through local reconstruction.

Inventive Principle:
Principle #2Taking out (Extraction)

3Measurement precision

If all reference samples are used for MIP prediction, then prediction accuracy is improved, but computational complexity increases

Engineering Contradiction:
Improveprediction accuracyVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts and uses only the most relevant reference samples (top and left boundaries) for MIP prediction, discarding less useful reference samples. This selective approach maintains prediction accuracy for typical cases while reducing computational complexity by processing fewer reference samples.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies different reference sample selection strategies based on local block characteristics and prediction directions. By adapting which reference samples to use based on local requirements rather than uniformly using all available samples, the system optimizes the balance between prediction accuracy and computational complexity for different regions.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP4694116A1MIP-based image encoding/decoding method, method for transmitting bitstream, and recording medium having bitstream stored therein
Publication Date: 2026.02.11 LG ELECTRONICS INC
  • EP4694116A1 patent drawingFigure 1
  • EP4694116A1 patent drawingFigure 2
  • EP4694116A1 patent drawingFigure 3

AI summary

An image encoding/decoding method and apparatus are provided. The image decoding method according to the present disclosure comprises the steps of: determining an intra prediction mode of the current block; and generating a prediction block of the current block by performing intra prediction on the basis of the determined intra prediction mode, wherein the intra prediction may be restricted to be performed by using only a reference sample within a predetermined range from among available reference samples of the current block.