Packing Strategy Signaling for 3D Point Cloud Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current point cloud compression methods in three-dimensional graphics fix patch orientation and position, leading to inefficient packing and compression, as they do not allow for flexible orientation or adaptive encoding of patch information.
Innovation Solution
A method for signaling patch orientation in 2D canvas images, enabling flexible orientation of geometry and texture patches through adaptive and fixed packing strategies, where orientation information is encoded and transmitted to allow for rotation and mirroring, optimizing packing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed patch orientation and position are used in point cloud compression, then the encoding process is simple, but the packing efficiency and compression performance deteriorate
Solution Approach 1:
The patent introduces dynamic patch orientation by allowing patches to be rotated to different orientations (0, 90, 180, 270 degrees) based on their content characteristics, transforming the static fixed-orientation approach into a dynamic adaptive approach that optimizes packing efficiency while maintaining manageable encoding complexity through systematic rotation rules
Solution Approach 2:
The patent changes the orientation parameter of patches from a fixed state to a variable state, where each patch can be rotated to one of four standard orientations. This parameter change enables better packing density and compression performance while the rotation angles are constrained to discrete values to control encoding complexity
2Device complexity
If the same background filling process is used for both geometry and texture, then the encoding process is simplified, but the compression performance deteriorates
Solution Approach 1:
The patent applies different background filling strategies to different patch types (geometry vs. texture) based on their local characteristics. Geometry patches use one filling approach while texture patches use another, allowing each type to be optimized for its specific data characteristics and improving overall compression performance without significantly increasing complexity
3Productivity
If flexible patch orientation with rotation and mirroring is implemented, then packing efficiency improves, but the signaling and encoding complexity increases
Solution Approach 1:
The patent uses asymmetric orientation assignment where patches are rotated to specific orientations based on their content characteristics rather than applying symmetric transformations uniformly. This asymmetric approach optimizes packing efficiency while the systematic selection of rotation angles (0, 90, 180, 270 degrees) keeps the signaling complexity manageable through constrained discrete choices
Solution Approach 2:
The patent performs preliminary orientation optimization during the patch packing stage by pre-determining the optimal orientation for each patch before final encoding. This preliminary action allows the system to achieve better packing efficiency upfront, reducing the need for complex post-processing and minimizing the overall encoding complexity
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for signaling the packing method of 2D patches in canvas images, used in coding of 3D point clouds is described herein. The method utilizes a signal representation to indicate the orientation method to be used by the patches. Furthermore, an optimal method of coding the orientation information of each patch is described.