Recursive Prediction Unit Partitioning in Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face challenges in efficiently managing the partitioning of video data into coding units, prediction units, and transform units, which affects compression efficiency and bandwidth utilization, especially as the demand for digital video increases with more connected devices.
Innovation Solution
The method involves recursively splitting coding units into prediction units and transform units, using various tree structures such as quad trees, binary trees, and ternary trees to optimize partitioning, and encoding these units into a bitstream with syntax elements indicating splits and prediction modes, allowing for flexible and efficient encoding and decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If video data is partitioned into coding units, prediction units, and transform units using traditional methods, then the coding structure is simple, but compression efficiency is insufficient
Solution Approach 1:
The patent applies segmentation by dividing video data into multiple hierarchical levels: coding tree units (CTUs) are split into coding units (CUs), which are further divided into prediction units (PUs) and transform units (TUs). This multi-level segmentation enables more precise compression by allowing different processing strategies at each level, directly improving compression efficiency while managing complexity through structured organization.
Solution Approach 2:
The patent implements nesting through hierarchical partitioning structures where CTUs contain CUs, which contain PUs and TUs. These nested structures allow progressive refinement of video data representation, with each level providing more detailed control over specific regions. This nesting enables efficient compression by applying appropriate processing at each hierarchical level without requiring complete re-processing of entire blocks.
2Productivity
If recursive splitting is applied to optimize partitioning, then compression efficiency improves, but processing complexity increases
Solution Approach 1:
The patent applies dynamics through adaptive recursive splitting where the partitioning process dynamically adjusts based on content characteristics. The encoder recursively divides CTUs into CUs and further into PUs and TUs based on local video content properties, allowing the structure to adapt to different regions. This dynamic approach improves compression efficiency by matching the partitioning structure to the actual content variations while managing processing complexity through systematic recursive algorithms.
Solution Approach 2:
The patent utilizes parameter changes by varying partitioning depth, split directions, and unit sizes based on content analysis. The recursive splitting process changes parameters such as block size and partitioning level adaptively, allowing optimal compression for different content types. This parameter adaptation improves compression efficiency without requiring completely different processing approaches for each content type.
3Adaptability or versatility
If multiple tree structures (quad trees, binary trees, ternary trees) are used for partitioning, then adaptability to different video content improves, but device complexity increases
Solution Approach 1:
The patent implements universality by designing a unified recursive partitioning framework that can accommodate multiple tree structures (quad trees, binary trees, ternary trees) within the same CTU-CU-PU-TU hierarchy. This multi-functional approach allows the system to select appropriate partitioning strategies based on content characteristics while maintaining a consistent overall structure. The universal framework improves adaptability to different video content types without requiring separate processing systems for each tree structure.
Data Source
AI summary
A mechanism for processing video data implemented by a video coding apparatus is disclosed. The mechanism determines to split a coding tree unit (CTU) into one or more coding units (CUs). One or more of the CUs may be one or more prediction tree units (PTUs). The mechanism further determines to recursively split the CUs into prediction units (PUs). The method performs a conversion between a visual media data and a bitstream based on the PUs.