Video Block Partitioning With Quad-Binary Trees for Intra Coding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video coding standards, such as ITU-T H.265 and JEM, lack flexibility in partitioning video blocks for improved compression efficiency, particularly in handling intra prediction types, leading to inefficiencies in data representation and potential artifacts.

Innovation Solution

Implementing a quad tree binary tree partitioning structure for both luma and chroma components, allowing for independent partitioning and further partitioning of nodes associated with intra prediction types, enhancing the flexibility and efficiency of video block partitioning.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a fixed partitioning structure is used for video blocks, then the coding process is simple, but the compression efficiency is insufficient and flexibility is limited

Engineering Contradiction:
Improvepartitioning flexibilityVSAvoidpartitioning structure complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The video block partitioning is divided into multiple stages using different tree structures. First, a quadtree structure divides the block into four quadrants, and then binary trees can further split selected quadrants. This segmentation allows flexible adaptation to different video content characteristics while maintaining a systematic approach to partitioning.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The partitioning structure transitions from a static fixed division to a dynamic multi-stage process where the quadtree and binary tree structures can be selectively applied based on video content characteristics. This allows the partitioning to adapt dynamically to different prediction types and content requirements.

Inventive Principle:
Principle #15Dynamics

2Productivity

If intra prediction types are handled with the same partitioning as inter prediction, then the processing is uniform, but the compression efficiency is reduced

Engineering Contradiction:
Improvecompression efficiencyVSAvoidprocessing complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

Different partitioning strategies are applied to different prediction types within the same video block. Intra prediction types receive specialized handling with additional binary tree partitioning options, while inter prediction types use the standard quadtree partitioning. This local differentiation optimizes compression efficiency for each prediction type without requiring complete system redesign.

Inventive Principle:
Principle #3Local quality

3Manufacturing precision

If video blocks are not partitioned finely enough, then the processing is faster, but the data representation accuracy is insufficient and artifacts appear

Engineering Contradiction:
Improvedata representation accuracyVSAvoidprocessing time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The quadtree structure performs an initial coarse partitioning of the video block into four quadrants before applying more detailed binary tree partitions. This preliminary action establishes a framework that guides subsequent finer partitioning, ensuring that detailed data representation is achieved efficiently without unnecessary processing steps.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3804315B1Systems and methods for partitioning video blocks in an inter prediction slice of video data
Publication Date: 2025.11.26 SHARP KK
  • EP3804315B1 patent drawingFigure 1
  • EP3804315B1 patent drawingFigure 2
  • EP3804315B1 patent drawingFigure 3

AI summary

A method of partitioning video data for video coding is disclosed. According to the method, a video block including sample values for a first component of video data and a second component of video data is received; the sample values for the first component of video data and the second component of video data is partitioned according to a first quad tree binary tree partitioning structure; and for samples values of the first component of video data included in nodes of the first quad tree binary tree associated with an intra prediction type, the samples is further partitioned according to a second quad tree binary tree.