Point Cloud Tile Bounding Box Constraints for G-PCC
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional point cloud coding techniques lack efficient constraints for bounding boxes of tiles in geometry-based Point Cloud Compression (G-PCC), leading to suboptimal processing efficiency and spatial random access in point cloud data compression.
Innovation Solution
Introducing constraints that ensure a tile's bounding box contains all points within its slices and excludes points from outside slices, along with aligning tile_id_bits with slice_tag_bits for precise spatial region definition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional point cloud coding techniques are used without bounding box constraints for tiles, then the coding standard is simpler to implement, but the processing efficiency and spatial random access performance deteriorate
Solution Approach 1:
The point cloud is divided into tiles, each with its own bounding box and slice set. This segmentation allows independent processing of spatial regions, improving processing efficiency and enabling spatial random access while maintaining manageable complexity through structured division
Solution Approach 2:
The bounding box of each tile is pre-calculated to contain all points in the tile's slices before encoding. This preliminary action ensures that spatial constraints are established in advance, enabling efficient processing and random access without adding complexity during the encoding process itself
2Measurement precision
If tile bounding boxes are constrained to contain all points in their slices, then spatial region accuracy is improved, but the complexity of determining and enforcing these constraints increases
Solution Approach 1:
The bounding box for each tile is pre-determined to encompass all points in the tile's slices during the preparation phase. This preliminary calculation ensures spatial accuracy is achieved before encoding, avoiding complex constraint enforcement during the coding process
Solution Approach 2:
The bounding box constraint is automatically satisfied by the tile construction process itself, as the bounding box is defined to contain all points in the tile's slices. This self-service approach ensures spatial accuracy without requiring additional complex verification or enforcement mechanisms
3Measurement precision
If tile_id_bits is aligned with slice_tag_bits, then spatial region definition precision is improved, but the bitstream structure becomes more complex
Solution Approach 1:
The alignment of tile_id_bits with slice_tag_bits creates a universal coding structure where the same bit depth serves both tile identification and slice tagging functions. This multi-functionality approach improves spatial region definition precision while avoiding additional complexity by reusing existing bit structures
Data Source
AI summary
Embodiments of the present disclosure provide a solution for point cloud coding. A method for point cloud coding is proposed. The method comprises: performing a conversion between a point cloud frame of a point cloud sequence and a bitstream of the point cloud sequence, a tile in the point cloud frame comprising a set of slices, a bounding box of the tile comprising at least a part of points in the set of slices, the at least a part of the points being obtainable from the set of slices.