Rectangular Tile Group Signaling With Derived Partition Locations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing HEVC and VVC video coding standards face inefficiencies in signaling rectangular tile groups, particularly in complying with constraints such as ensuring each tile has its entire left and top boundaries at picture boundaries or previously decoded tiles, which leads to suboptimal bitstream processing.
Innovation Solution
A method for deriving tile group locations based on previously signaled and derived tile group positions, rather than explicitly specifying both top-left and bottom-right positions, using syntax elements to ensure compliance with rectangular tile group requirements, thereby optimizing bitstream efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If both top-left and bottom-right positions of tile groups are explicitly signaled, then complete position information is provided, but bitstream length increases and processing complexity increases
Solution Approach 1:
The patent extracts only the necessary position information by removing the redundant top-left position signaling. The bottom-right position is derived by adding the tile group dimensions to the top-left position, eliminating the need to signal both positions explicitly and thus reducing bitstream length and processing complexity.
Solution Approach 2:
The top-left position is derived self-service by using the bottom-right position and tile group dimensions. The system calculates the top-left position as bottom-right minus dimensions, eliminating the need for explicit top-left signaling and reducing the overall complexity of position information processing.
2Measurement precision
If both top-left and bottom-right positions are explicitly signaled, then complete tile group location is provided, but bit cost increases
Solution Approach 1:
The patent removes the redundant top-left position signaling from the bitstream. By deriving the top-left position from the bottom-right position and tile group dimensions, the system maintains complete location information while reducing the number of bits required to signal tile group positions.
Solution Approach 2:
The patent changes the parameter representation from signaling both top-left and bottom-right positions to signaling only the bottom-right position and dimensions. This parameter transformation maintains the completeness of location information while reducing the bit cost through more efficient encoding.
3Reliability
If tile group constraints are enforced through explicit signaling, then compliance is ensured, but processing time increases
Solution Approach 1:
The patent performs preliminary action by deriving the top-left position from the bottom-right position and tile group dimensions during the decoding process. This pre-calculation ensures that the tile group constraints are automatically satisfied without requiring additional verification steps, thus maintaining reliability while reducing processing time.
Solution Approach 2:
The system uses self-service by automatically ensuring constraint compliance through the mathematical relationship between bottom-right position, dimensions, and top-left position. The derivation process inherently guarantees that tile groups satisfy the required constraints, eliminating the need for separate validation and reducing processing time.
Data Source
AI summary
A method for decoding a picture from a bitstream. The method includes decoding a tile partition structure from one or more syntax elements in the bitstream. The method includes determining a number of tiles N in the picture. The method includes decoding a number of partitions in the picture, wherein each partition comprises an integer number of tiles. The method includes deriving a size and/or location for a current partition, wherein deriving the size and/or location for the current partition comprises: decoding one or more location syntax elements for the current partition from the bitstream, deriving a top-left position of the current partition as a location of a first tile in raster scan order that is not included in one or more previously derived partition, and deriving the size and/or the location of the current partition from a value of the one or more location syntax element together with a size and/or location of the one or more previously derived partition. The method includes using the derived size and/or location to decode the picture.