Occurrence-Based Intra Coding Using Neighbor Histograms

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video coding systems face inefficiencies in predicting intra prediction modes, leading to suboptimal compression and decoding performance.

Innovation Solution

A device or method that determines intra prediction modes by analyzing neighboring blocks, updating histograms based on sample numbers, and using blending weights to refine intra prediction mode occurrences for improved encoding and decoding.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional intra prediction mode selection is used, then encoding complexity is reduced, but prediction accuracy and compression efficiency deteriorate

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

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing histograms of intra prediction mode occurrences from previously decoded neighboring blocks. These pre-computed histograms are then directly used to guide the selection of intra prediction modes for the current block, eliminating the need for complex real-time analysis during encoding while improving prediction accuracy through data-driven mode selection.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements self-service by having the decoder perform the histogram analysis and intra prediction mode selection that would traditionally require encoder involvement. The decoder uses the pre-computed histograms to automatically determine the most probable intra prediction modes for current blocks, reducing encoder complexity while maintaining prediction accuracy through autonomous decoder-side decision-making.

Inventive Principle:
Principle #25Self-service

2Measurement precision

If histogram updating based on multiple neighboring blocks is performed, then intra prediction mode prediction accuracy is improved, but computational complexity increases

Engineering Contradiction:
Improveintra prediction mode prediction accuracyVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSPower

Solution Approach 1:

The patent applies the extraction principle by selectively using only the necessary histogram data from neighboring blocks that are already decoded and available. Instead of analyzing all possible neighboring blocks or performing exhaustive histogram computations, the method extracts and utilizes pre-computed histogram information from relevant neighboring blocks, reducing computational complexity while maintaining prediction accuracy.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent uses preliminary action by pre-computing histograms during the decoding of neighboring blocks before they are needed for the current block's prediction. This advance preparation eliminates the need for complex real-time histogram analysis during the critical path of encoding/decoding the current block, reducing computational complexity while preserving accurate prediction mode selection.

Inventive Principle:
Principle #10Preliminary action

3Manufacturing precision

If blending weights are calculated for multiple intra prediction modes, then prediction quality is enhanced, but encoding overhead increases

Engineering Contradiction:
Improveprediction qualityVSAvoidencoding overhead
Core Design Contradiction:
Manufacturing precisionVSLoss of information

Solution Approach 1:

The patent applies parameter changes by dynamically adjusting the number of blending weights calculated based on the specific characteristics of the current block and its neighboring blocks. Instead of always calculating blending weights for all possible intra prediction modes, the method adapts the parameter of how many modes receive blending weight calculations, enhancing prediction quality where needed while minimizing encoding overhead in simpler cases.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent implements partial action by calculating blending weights for only the most probable intra prediction modes identified from the histogram analysis, rather than computing weights for all possible modes. This selective approach enhances prediction quality for the dominant modes while avoiding the excessive computational and encoding overhead that would result from processing all modes equally.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4633147A1Occurrence-based intra coding (OBIC)
Publication Date: 2025.10.15 INTERDIGITAL CE PATENT HOLDINGS SAS
  • EP4633147A1 patent drawingFigure 1A
  • EP4633147A1 patent drawingFigure 1B
  • EP4633147A1 patent drawingFigure 1C

AI summary

A device determine that a neighboring block associated with a current block is associated with a plurality of intra prediction modes. The device may obtain a sample number associated with a first intra prediction mode. The device may update a histogram of intra prediction mode occurrence associated with the current block based on the sample number associated with the first intra prediction mode and the first intra prediction mode. The device may encode and/or decode the current block based on the histogram of intra prediction mode occurrence for the current block. The neighboring block may include a plurality of partitions. The device may identify a first partition of the neighboring block associated with the first intra prediction mode and a second partition associated with the second the first intra prediction mode, and update the histogram of intra prediction mode occurrence based on the sample numbers of the two partitions.