3D Point Cloud Patch Mapping for Flexible 2D Canvas Packing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current point cloud compression methods using video encoders face limitations such as fixed patch orientation, co-location of geometry and texture patches, and uniform background filling, leading to sub-optimal coding efficiency.
Innovation Solution
The method involves flexible mapping of 3D patch surfaces into 2D canvases by independently placing geometry and attribute patches, allowing for different orientations, luminance adjustments, and using safeguard distances and background filling techniques to optimize packing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If patches are co-located in the same 2D canvas for geometry and texture, then alignment is simplified, but coding efficiency decreases due to sub-optimal packing structure
Solution Approach 1:
The patent divides the 2D canvas into separate canvases for geometry patches and texture patches. This segmentation allows each canvas to be optimized independently for its specific data type, improving packing efficiency while maintaining the alignment relationship through separate but coordinated placement strategies.
Solution Approach 2:
The patent transitions from a single 2D canvas to multiple 2D canvases, adding the dimension of canvas separation. This allows geometry and texture patches to be placed in different spatial arrangements optimized for their respective characteristics while still maintaining their correspondence through metadata.
2Device complexity
If patch orientation is fixed, then processing is simplified, but space utilization decreases leading to sub-optimal packing
Solution Approach 1:
The patent introduces dynamic orientation capabilities where patches can be rotated to different orientations (e.g., 0, 90, 180, 270 degrees) based on their content and the available space in the canvas. This dynamic adjustment optimizes space utilization while maintaining manageable processing complexity through predefined orientation options.
Solution Approach 2:
The patent changes the orientation parameter of patches from a fixed value to a variable that can take multiple discrete values. This allows the system to select the optimal orientation for each patch based on space availability and content characteristics, improving packing efficiency.
3Ease of operation
If uniform background filling is applied to both geometry and texture, then processing is consistent, but coding efficiency decreases due to different characteristics of geometry and texture data
Solution Approach 1:
The patent applies different background filling strategies to geometry and texture patches based on their local characteristics. Geometry patches may use one filling approach while texture patches use another, optimizing the coding efficiency for each data type while maintaining overall processing consistency through a unified framework.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods for mapping 3D point cloud data into 2D surfaces are described herein. The methods utilize 3D surface patches to represent point clouds and perform flexible mapping of 3D patch surface data into 2D canvas images. Patches representing geometry and patches representing attributes such as textures are placed in different canvases, where the placement of each patch is done independently for geometry and texture, that is, geometry and texture patches do not need to be co-located, as in conventional point cloud mapping. Furthermore, methods include transformations of the 3D patch when placing it into the 2D canvas, for more efficient packing.