Video Block Binary Tree Partitioning for Higher Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current HEVC video coding systems rely solely on quadtree partitioning for CTU, CU, and PU, which limits coding efficiency and flexibility, especially in meeting the increasing demands of video storage and transmission.
Innovation Solution
A method employing binary tree partitioning processes, which can be applied recursively to video data blocks, allowing for symmetric and asymmetric partitions, and combining with quadtree partitioning to enhance coding efficiency and adaptability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If quadtree partitioning is used for CTU, CU, and PU, then the partitioning process is simple and systematic, but coding efficiency is limited and adaptability to varying video content characteristics is poor
Solution Approach 1:
The partitioning process is segmented into two distinct stages: quadtree partitioning for coarse block division and binary tree partitioning for fine-grained subdivision. This two-stage segmentation allows the system to first establish a systematic partitioning framework with quadtree, then enhance adaptability through binary tree partitioning that can create asymmetric and varied block shapes tailored to local video characteristics.
Solution Approach 2:
The binary tree partitioning introduces asymmetric partitioning modes that allow blocks to be divided unevenly based on content characteristics, unlike the symmetric quadtree partitioning. This asymmetry enables better adaptation to varying video content by creating partition patterns that match edge directions and content boundaries more effectively.
2Productivity
If quadtree partitioning is used, then the partitioning structure is regular and easy to implement, but coding efficiency and prediction accuracy are insufficient
Solution Approach 1:
The partitioning structure transitions from the static, fixed quadtree pattern to a dynamic hybrid structure where binary tree partitioning can be selectively applied based on content characteristics. This dynamic approach allows the system to adapt partition patterns to local variations in video content, improving coding efficiency by matching partition boundaries to content boundaries.
Solution Approach 2:
The invention changes the partitioning parameters by introducing binary tree partitioning with different splitting modes (horizontal, vertical, asymmetric) alongside quadtree partitioning. This parameter expansion allows the system to vary partition patterns according to content needs, improving prediction accuracy and coding efficiency without requiring complete restructuring of the partitioning framework.
3Adaptability or versatility
If minimum CU size is set to 8×8, then the granularity for switching prediction types is fixed, but flexibility in adapting to different block sizes and characteristics is reduced
Solution Approach 1:
The partitioning control is segmented into multiple decision points where binary tree partitioning can create various block sizes beyond the fixed 8×8 minimum. This segmentation allows different prediction types to be switched at multiple granularity levels, providing flexibility in adapting to different block sizes and characteristics while maintaining systematic control through the two-stage partitioning framework.
Data Source
AI summary
An apparatus and method for video encoding comprising one or more electronic devices configured to: receive input data associated with a coding unit (CU) of video data, determine a binary tree partitioning structure corresponding to a block partitioning process including a binary tree partitioning process for the CU, wherein the binary tree partitioning structure represents partitioning the CU into a plurality of transform units (TUs), and when the binary tree partitioning process decides to apply binary tree partition to one given CU, said one given CU is always split into two Tus, and apply an encoding process comprising transform process to the CU by applying the encoding process at a level corresponding to the TUs.