Content Adaptive Video Frame Slicing and Non-Uniform Access Unit Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video compression standards face inefficiencies in coding due to rigid slice structures and uniform coding types across entire video frames, which are inadequate for handling global motion and scene changes, leading to suboptimal compression and potential decoding errors.
Innovation Solution
Implementing content adaptive video frame slicing and non-uniform access unit coding using flexible macroblock ordering (FMO) within a single video access unit, allowing for different slice coding types and adaptive partitioning based on global motion detection and scene changes, such as camera pans or composite scenes with semantically different segments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If uniform coding types are used across the entire video frame, then decoding simplicity is maintained, but coding efficiency deteriorates during global motion and scene changes
Solution Approach 1:
The video frame is divided into multiple slices with different coding types (I-slice, P-slice, B-slice) within a single access unit. This segmentation allows different regions to use optimal coding types based on local content characteristics, improving coding efficiency while maintaining manageable complexity through structured partitioning.
Solution Approach 2:
Different slices within the same access unit are assigned different coding types based on local scene characteristics. Regions with global motion or scene changes use I-slices for error resilience, while stable regions use P or B slices for compression efficiency, achieving local optimization without requiring complete frame-level uniformity.
2Adaptability or versatility
If rigid slice structures are used, then decoding process is simplified, but adaptability to global motion and scene changes deteriorates
Solution Approach 1:
The slice structure is made dynamic by allowing different coding types within a single access unit based on detected global motion and scene changes. The encoder adaptively determines slice boundaries and coding types according to content characteristics, providing flexibility while maintaining a regular slice-based structure that decoders can handle systematically.
Solution Approach 2:
Coding parameters such as slice coding type (I, P, B) are changed adaptively based on scene analysis. When global motion or scene changes are detected in specific regions, the coding type parameter is modified for those slices, allowing the system to adapt to content variations without fundamentally changing the slice structure itself.
3Productivity
If different slice coding types are used within a single VAU, then coding efficiency is improved, but decoder complexity increases
Solution Approach 1:
The bitstream is segmented into slice-specific data structures with clear headers indicating coding type. Each slice is independently decoded with its appropriate coding type, allowing the decoder to handle multiple types through systematic segmentation rather than requiring complex unified processing logic.
Solution Approach 2:
The decoder is designed with universal processing capabilities that can handle multiple slice coding types (I, P, B) within a single access unit using a unified decoding framework. The same decoder architecture processes all slice types by interpreting slice header information, avoiding the need for separate decoders for each coding type.
Data Source
AI summary
Techniques for content adaptive video frame slicing and non-uniform access unit coding for improved coding efficiency are provided. An encoder and decoder are disclosed to process (encode or decode) a single non-uniform video access unit (VAU) employing flexible macroblock ordering (FMO) in conjunction with different slice coding types in response to global motion detection of a camera pan or a scroll within the single VAU.