Intra Prediction Search Range Constraint for Parallel CTU Processing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video coding technologies face challenges in efficiently processing multiple coding tree units (CTUs) in parallel, which hinders coding efficiency and quality.

Innovation Solution

The method involves determining either a reference block or an intra prediction mode for a current block within a CTU, constrained to allow parallel processing of adjacent CTUs, and using this information to reconstruct the current block.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If intra prediction uses reference blocks or templates from adjacent CTUs to improve prediction accuracy, then video quality improves, but parallel processing capability deteriorates

Engineering Contradiction:
Improveprediction accuracyVSAvoidparallel processing capability
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The patent extracts only the necessary reference samples from the current CTU that do not overlap with adjacent CTUs. By taking out only the non-overlapping portion of reference blocks and templates, the patent maintains prediction accuracy while eliminating the harmful dependency on adjacent CTUs, thus enabling parallel processing.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the reference search area into overlapping and non-overlapping regions relative to adjacent CTUs. By segmenting the template area and reference block search range, the patent identifies which portions can be safely used for parallel processing and which would create dependencies, allowing selective use of reference data.

Inventive Principle:
Principle #1Segmentation

2Manufacturing precision

If the search range for reference blocks is expanded to improve prediction quality, then video quality improves, but processing complexity increases

Engineering Contradiction:
Improveprediction qualityVSAvoidprocessing complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent applies different constraints to different regions of the search space. Instead of uniformly limiting or expanding the search range, the patent locally adjusts the search area based on the spatial relationship with adjacent CTUs, allowing maximum search range in safe regions while imposing constraints only where parallel processing would be affected.

Inventive Principle:
Principle #3Local quality

3Productivity

If template-based intra prediction is applied to all blocks to improve compression efficiency, then coding efficiency improves, but processing time increases

Engineering Contradiction:
Improvecoding efficiencyVSAvoidprocessing time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent applies template-based intra prediction selectively rather than universally. By applying the technique only to blocks where it can be computed independently (using constrained reference areas), the patent achieves partial application that maintains coding efficiency benefits while avoiding the full processing overhead that would prevent parallel execution.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS20250126270A1Restricted region of intra prediction to support parallel processing
Publication Date: 2025.04.17 TENCENT AMERICA LLC
  • US20250126270A1 patent drawing
  • US20250126270A1 patent drawing
  • US20250126270A1 patent drawing

AI summary

Aspects of the disclosure includes methods and apparatuses for video decoding and video encoding and a method of processing visual media data. A current block in a first coding tree unit (CTU) is predicted using samples in a current picture. The first CTU and a second CTU are processed in parallel. The method for video decoding includes determining one of (i) a reference block of the current block in an allowed search range that is constrained for parallel processing of the first CTU and the second CTU in the current picture and (ii) an intra prediction mode of the current block based on a template adjacent to the current block that is constrained for parallel processing of the first CTU and the second CTU, and reconstructing the current block using the one of (i) the reference block and (ii) the intra prediction mode.