Video Decoder Block Signaling for Extended Ternary Partitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards face inefficiencies in partitioning video blocks, leading to suboptimal bandwidth usage and compression performance, particularly with non-power-of-two sized sub-blocks.
Innovation Solution
Implementing an Extended Ternary-Tree (ETT) partitioning scheme that allows for sub-blocks with non-power-of-two dimensions, enabling more flexible and efficient coding by partitioning parent blocks into three sub-blocks with specific width and height ratios, and incorporating this partitioning into the bitstream signaling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional binary tree partitioning is used, then the partitioning structure is simple and easy to implement, but the coding efficiency is suboptimal and cannot adapt to non-power-of-two block sizes
Solution Approach 1:
The patent extends the traditional binary tree partitioning by introducing a ternary tree structure that divides a parent block into three sub-blocks instead of two. This segmentation approach allows for more flexible partitioning patterns while maintaining a systematic tree-based organization, thereby improving coding efficiency without excessive complexity increase
Solution Approach 2:
The patent modifies the partitioning parameters by allowing non-power-of-two block sizes in the ternary tree structure. Traditional video coding standards restrict blocks to power-of-two dimensions, but this invention changes that parameter constraint to enable more adaptive partitioning that better matches natural image structures and improves compression performance
2Adaptability or versatility
If non-power-of-two sized sub-blocks are introduced, then the partitioning flexibility is improved, but the compatibility with existing video coding standards deteriorates
Solution Approach 1:
The patent introduces asymmetric partitioning through the ternary tree structure where the parent block is divided into three sub-blocks of potentially different sizes and shapes. This asymmetric division provides greater adaptability to various block size requirements while the overall tree structure maintains systematic organization that can be integrated with existing coding frameworks
3Productivity
If extended ternary-tree partitioning is implemented, then the compression performance is improved, but the bandwidth requirements for signaling partition information increase
Solution Approach 1:
The patent extracts and separates the partitioning information signaling from the main video data stream, allowing for efficient encoding of the ternary tree structure. By dedicating specific signaling mechanisms for partition information, the system achieves better compression performance while managing the additional bitstream requirements through structured and compact representation of partition data
Data Source
AI summary
A mechanism for processing video data is disclosed. A parent block is partitioned with an Extended Ternary-Tree (ETT) partition to create three sub-blocks. At least one of the sub-blocks includes a side with a measurement that is not a power of two. A conversion is performed between a visual media data and a bitstream based on the sub-blocks.