Video Block Partitioning With Nested QTBT for Artifact-Aware Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards like ITU-T H.265 and JEM face challenges in efficiently partitioning video blocks for improved compression and reduced artifacts, particularly in intra and inter prediction techniques, transform techniques, and entropy coding.
Innovation Solution
The proposed method involves partitioning video blocks using a combination of quadtree and binary tree structures, allowing for independent partitioning of luma and chroma components, and applying advanced filtering techniques like adaptive loop filtering to minimize artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If video blocks are partitioned using existing standards like ITU-T H.265 and JEM, then video compression is achieved, but visual artifacts and compression efficiency are not optimized
Solution Approach 1:
The video block partitioning process is segmented into multiple independent tree structures: a first quadtree binary tree for initial partitioning, and a second quadtree binary tree for further partitioning of specific nodes. This multi-level segmentation allows for more granular and adaptive block division, improving compression efficiency while reducing artifacts by better matching content characteristics.
Solution Approach 2:
The patent implements a nested partitioning structure where the second quadtree binary tree is applied within and builds upon the partitioning results of the first quadtree binary tree. Nodes from the first tree that meet specific criteria are selected as root nodes for the second tree, creating a nested hierarchy that enables progressive refinement of partitioning without redundant processing.
2Productivity
If a single quadtree partitioning structure is used, then implementation is simple, but compression efficiency and artifact reduction are limited
Solution Approach 1:
The patent introduces dynamic adaptability by applying the second quadtree binary tree selectively only to nodes that meet specific criteria (intra prediction type and/or inter prediction type with specific reference picture list). This dynamic approach allows the system to increase complexity only where necessary for improved compression, rather than uniformly across all blocks.
Solution Approach 2:
Different partitioning strategies are applied to different regions of the video block based on local characteristics. The second quadtree binary tree is applied locally to specific nodes requiring further partitioning, while other nodes maintain the partitioning from the first tree. This local quality approach optimizes compression efficiency for specific regions without unnecessarily increasing overall complexity.
3Manufacturing precision
If luma and chroma components are partitioned together, then processing is simplified, but component-specific optimization is lost
Solution Approach 1:
The patent applies the second quadtree binary tree independently to luma and chroma components based on their specific prediction types and characteristics. Each component can have different nodes selected for further partitioning, allowing component-specific optimization while maintaining a unified overall framework that prevents excessive complexity.
Solution Approach 2:
The partitioning process is segmented into component-specific stages where luma and chroma components are processed separately through the second quadtree binary tree based on their individual requirements. This segmentation enables precise, component-optimized partitioning while the modular structure keeps the overall process manageable.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A video coding device may be configured to perform video coding comprising: receiving a video block including sample values for a first component of video data and a second component of video data; partitioning the sample values for the first component of video data and the second component of video data according to a first quad tree binary tree (QTBT) 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, further partitioning the samples according to a second quad tree binary tree.