Rectangular Tile Grouping for Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing video coding standards, such as HEVC, face challenges in interactive performance and coding complexity due to the lack of effective tile and slice partitioning, which limits parallel processing and region-of-interest functionality.
Innovation Solution
The introduction of rectangular tile groups, which replace traditional tiles and slices, allows for non-overlapping rectangular divisions of video pictures, enabling flexible grouping and filtering operations across boundaries, and signaling flags to control filtering processes, improving encoding and decoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional tiles and slices are used for partitioning video pictures, then parallel processing capability is limited, but coding complexity increases and interactive performance deteriorates
Solution Approach 1:
The video picture is divided into multiple independent tile groups, where each tile group contains one or more tiles. This segmentation allows parallel processing of different tile groups while reducing the complexity of inter-tile dependencies. The rectangular shape constraint and independent decoding capability of each tile group enable efficient parallel processing without requiring complex coordination between processing units.
Solution Approach 2:
The patent introduces dynamic control mechanisms through flags (e.g., loop_filter_across_tile_groups_enabled_flag) that allow the filtering operation to adaptively switch between being performed across tile group boundaries and being performed only within individual tile groups. This dynamic approach optimizes the balance between processing efficiency and output quality based on content characteristics.
2Manufacturing precision
If filtering operations are performed across tile boundaries, then output quality improves, but processing complexity and time increase
Solution Approach 1:
The patent applies different filtering strategies to different regions of the video picture based on their characteristics. By allowing filtering across tile group boundaries only where beneficial (controlled by the loop_filter_across_tile_groups_enabled_flag), the system maintains high output quality in regions that benefit from cross-boundary filtering while avoiding unnecessary processing in regions where it would not improve quality.
Solution Approach 2:
The patent changes the parameter of filtering scope by introducing the loop_filter_across_tile_groups_enabled_flag, which can switch the filtering operation between two states: (1) filtering across tile group boundaries, and (2) filtering only within individual tile groups. This parameter change allows the system to adapt processing time and quality based on specific content requirements.
3Productivity
If rectangular tile groups with flexible grouping are introduced, then encoding efficiency improves, but device complexity increases
Solution Approach 1:
The patent creates a universal structure where tile groups serve multiple functions: they are basic units for parallel processing, rectangular regions for efficient memory access patterns, and controllable filtering boundaries. This multi-functionality improves encoding efficiency without requiring separate specialized structures for each function, thereby limiting the increase in device complexity.
Solution Approach 2:
The patent organizes tiles into rectangular tile groups, adding a grouping dimension on top of the existing tile structure. This dimensional organization enables more efficient memory access patterns and parallel processing layouts without fundamentally changing the underlying tile-based coding structure, thus improving encoding efficiency with minimal increase in device complexity.
Data Source
AI summary
Methods and apparatuses for processing video data include receiving input data associated with a current video picture, dividing the current video picture into non-overlapping rectangular tiles, grouping the tiles in the current video picture into tile groups, and encoding or decoding video data in the tile groups within the current video picture. According to one embodiment, each tile group is composed of an integer number of tiles, and shapes of all the tile groups are constrained to be rectangle. According to one embodiment, a flag is used to indicate whether one or more in-loop filtering operations are performed across tile group boundaries.