Intra Prediction Mode Encoding Using Candidate Lists

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The increase in intra-frame prediction modes in video encoding technologies leads to a reduction in encoding efficiency due to reduced statistical correlation between prediction modes and increased bit count for encoding prediction mode information.

Innovation Solution

An image predictive encoding method that generates a candidate prediction mode list for neighboring blocks, encodes a flag to indicate the presence of the optimum prediction mode, and uses an index or reassigns numbers to encode the prediction mode efficiently, even when the mode is not in the list.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the number of intra-frame prediction modes is increased to improve prediction accuracy, then the prediction accuracy is improved, but the encoding efficiency of prediction mode information deteriorates due to reduced statistical correlation and increased bit count

Engineering Contradiction:
Improveprediction accuracyVSAvoidencoding efficiency of prediction mode information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent applies preliminary action by pre-defining a limited set of candidate prediction modes (typically 3 modes: vertical, horizontal, and diagonal) that are most likely to be optimal for a given block type. This preliminary selection is performed before the actual encoding process, allowing the encoder to efficiently choose from a small, pre-optimized set of modes rather than searching through all possible modes, thereby maintaining high prediction accuracy while minimizing the bit count required for mode signaling.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent changes the parameter of candidate prediction mode selection by adapting the candidate modes based on block characteristics such as block size and block type (e.g., luma vs. chroma). Different block types are assigned different candidate mode sets, allowing the system to optimize prediction accuracy for each specific case while keeping the total number of candidates limited. This parameter-based adaptation enables the system to maintain encoding efficiency while improving prediction accuracy for diverse block types.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If more intra-frame prediction modes are provided to improve prediction accuracy, then the prediction accuracy is improved, but the bit count for encoding prediction mode information increases

Engineering Contradiction:
Improveprediction accuracyVSAvoidbit count for encoding prediction mode information
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent applies preliminary action by pre-defining a limited set of candidate prediction modes (typically 3 modes: vertical, horizontal, and diagonal) that are most likely to be optimal for a given block type. This preliminary selection is performed before the actual encoding process, allowing the encoder to efficiently choose from a small, pre-optimized set of modes rather than searching through all possible modes, thereby maintaining high prediction accuracy while minimizing the bit count required for mode signaling.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent changes the parameter of candidate prediction mode selection by adapting the candidate modes based on block characteristics such as block size and block type (e.g., luma vs. chroma). Different block types are assigned different candidate mode sets, allowing the system to optimize prediction accuracy for each specific case while keeping the total number of candidates limited. This parameter-based adaptation enables the system to maintain encoding efficiency while improving prediction accuracy for diverse block types.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If the number of intra-frame prediction modes is increased, then the prediction accuracy is improved, but the statistical correlation between prediction modes is reduced

Engineering Contradiction:
Improveprediction accuracyVSAvoidstatistical correlation between prediction modes
Core Design Contradiction:
Measurement precisionVSStability of the object's composition

Solution Approach 1:

The patent applies preliminary action by pre-defining a limited set of candidate prediction modes (typically 3 modes: vertical, horizontal, and diagonal) that are most likely to be optimal for a given block type. This preliminary selection is performed before the actual encoding process, allowing the encoder to efficiently choose from a small, pre-optimized set of modes rather than searching through all possible modes, thereby maintaining high prediction accuracy while minimizing the bit count required for mode signaling.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent changes the parameter of candidate prediction mode selection by adapting the candidate modes based on block characteristics such as block size and block type (e.g., luma vs. chroma). Different block types are assigned different candidate mode sets, allowing the system to optimize prediction accuracy for each specific case while keeping the total number of candidates limited. This parameter-based adaptation enables the system to maintain encoding efficiency while improving prediction accuracy for diverse block types.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4007280A1Intra prediction mode decoding
Publication Date: 2022.06.01 NTT DOCOMO INC
  • EP4007280A1 patent drawingFigure 1
  • EP4007280A1 patent drawingFigure 2
  • EP4007280A1 patent drawingFigure 3

AI summary

When a larger number of prediction modes are provided in generation of intra-frame predicted signals, mode information to identify an intra-frame prediction method of a target block is efficiently encoded. An image predictive encoding method includes: determination of an optimum prediction mode from a plurality of prediction methods for a pixel signal of a target block; a predicted signal is generated according to the mode; a residual signal is determined between the pixel signal of the target block and the predicted signal; the residual signal and the optimum prediction mode are encoded to generate a compressed signal; the compressed signal is restored; the restored signal is stored as a reproduced pixel signal. In the prediction mode encoding, a candidate prediction mode list is generated as one containing elements of optimum prediction modes of a plurality of previously-reproduced blocks neighboring the target block; a flag to indicate whether the list contains an element corresponding to the optimum prediction mode is encoded; an index to the corresponding element is encoded if the corresponding element is present in the list; the optimum prediction mode is encoded based on numbering the elements in the list, unless there is no corresponding element in the list.