Intra Prediction Reference Line Selection for Video Coding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video coding technologies face inefficiencies in intra prediction due to the use of a limited number of prediction directions, which results in increased bit usage for less likely directions, and motion vector prediction methods that do not fully leverage spatial and temporal correlations in video content.

Innovation Solution

The proposed solution involves determining a subset of reference lines for intra prediction based on the tangent value of the prediction angle and row number of the current block, allowing for a greater variety of prediction directions and improving the efficiency of intra prediction by reducing the bit usage for less likely directions, and enhancing motion vector prediction by utilizing surrounding metadata for more accurate motion compensation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a limited number of prediction directions are used in intra prediction, then the device complexity is reduced, but the compression efficiency deteriorates due to increased bit usage for less likely directions

Engineering Contradiction:
Improvenumber of prediction directionsVSAvoidbit usage efficiency
Core Design Contradiction:
Device complexityVSLoss of information

Solution Approach 1:

The patent segments the prediction directions into multiple groups or categories, allowing the encoder to select from a limited set of representative directions while using segmentation techniques to handle the diversity of prediction needs. This reduces the overall number of directions that must be explicitly coded while maintaining compression efficiency through structured organization of prediction options.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter representation by using tangent values of prediction angles instead of direct angle coding. This parameter transformation allows for more efficient entropy coding of prediction directions, reducing the average bit usage while maintaining the ability to represent a wide range of prediction angles through a compact set of discrete options.

Inventive Principle:
Principle #35Parameter changes

2Device complexity

If traditional motion vector prediction methods are used, then the device complexity is reduced, but the compression efficiency deteriorates due to insufficient exploitation of spatial and temporal correlations

Engineering Contradiction:
Improvemotion vector prediction methodVSAvoidcompression efficiency
Core Design Contradiction:
Device complexityVSLoss of information

Solution Approach 1:

The patent performs preliminary analysis of spatial and temporal correlations by examining surrounding metadata and reference frames before finalizing motion vector prediction. This preliminary action allows the system to identify and exploit correlation patterns in advance, improving prediction accuracy without requiring complex real-time processing during encoding.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces intermediary structures such as candidate motion vector lists and reference frame buffers that mediate between raw video data and final motion compensation. These intermediaries enable sophisticated exploitation of spatial and temporal correlations by providing structured access to historical and neighboring data, improving prediction efficiency while managing computational complexity.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP4042682B1Method and apparatus for video coding
Publication Date: 2026.03.04 TENCENT AMERICA LLC
  • EP4042682B1 patent drawingFigure 1A
  • EP4042682B1 patent drawingFigure 1B
  • EP4042682B1 patent drawingFigure 1C

AI summary

Aspects of the disclosure include methods, apparatuses, and non-transitory computer-readable storage mediums for video encoding/decoding. An apparatus includes processing circuitry that decodes prediction information of a current block in a current picture that is a part of a coded video sequence. The prediction information indicates one of a plurality of intra prediction directions for the current block. The processing circuitry determines a subset of a plurality of reference lines based on the one of the plurality of intra prediction directions indicated in the prediction information. The processing circuitry performs an intra prediction of the current block based on the determined subset of the plurality of reference lines. The processing circuitry reconstructs the current block based on the intra prediction of the current block.