Point Cloud Skeleton Extraction via Iterative Contraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for extracting a skeleton from a point cloud are inaccurate due to the presence of noise, exterior points, and missing data, as they rely heavily on normal vector information and assumptions about the object's shape, leading to low accuracy in reconstructing three-dimensional models.
Innovation Solution
A method that uses an iterative formula to contract point cloud sampling data, incorporating a regular term and expanding neighborhood radii, along with principal component analysis and k-Nearest Neighborhood algorithms, to improve the accuracy of skeleton extraction by reducing noise and enhancing the robustness of the skeleton branches generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional skeleton extraction methods relying on normal vector information and shape assumptions are used, then the processing can be performed with simple algorithms, but the extraction accuracy deteriorates due to noise, exterior points, and missing data in point cloud
Solution Approach 1:
The patent changes the fundamental parameters of the extraction approach by abandoning normal vector calculations and shape assumptions, instead using a direct point-to-skeleton distance minimization approach with iterative optimization. This parameter change allows accurate skeleton extraction from noisy, incomplete point clouds without relying on problematic intermediate representations.
Solution Approach 2:
The patent extracts only the essential geometric relationship (shortest distance from points to skeleton) while eliminating extraneous information (normal vectors, shape assumptions). This extraction principle focuses the algorithm on the core geometric truth that skeleton points minimize distance to surface points, regardless of noise or missing data.
2Reliability
If iterative contraction with regular term and expanding neighborhood radii is used, then the robustness of skeleton branches generation is improved, but the computation time increases
Solution Approach 1:
The patent employs periodic action through iterative contraction with expanding neighborhood radii. The algorithm repeatedly contracts the point cloud while systematically increasing the neighborhood radius in each iteration cycle. This periodic expansion allows the method to progressively capture skeleton structure at multiple scales, improving robustness against noise and missing data.
Solution Approach 2:
The patent performs preliminary action by pre-defining the iterative contraction framework and neighborhood expansion schedule before processing. The regular term and expansion strategy are established in advance, allowing the algorithm to systematically progress through iterations without adaptive decision-making during execution, which optimizes computational efficiency.
3Adaptability or versatility
If point cloud data with large-area missing is used for reconstruction, then the flexibility of scanning is maintained, but the ability to rebuild accurate three-dimensional models deteriorates
Solution Approach 1:
The patent converts the harm of missing data into a benefit by using the iterative contraction approach that naturally handles incomplete point clouds. Instead of requiring complete surfaces for normal vector calculation, the method uses available points to iteratively converge toward the skeleton, transforming data incompleteness from a obstacle into a feature that directly reveals the underlying structure.
Data Source
AI summary
A method for extracting a skeleton form a point cloud includes: obtaining inputted point cloud sampling data; contracting the point cloud using an iterative formula and obtaining skeleton branches, the iterative formula is:argminX∑i∈I∑j∈Jxi-qiθ(xj-qj)+R(X),whereinR(X)=∑i∈Iγi∑i′∈I\{i}θ(xi-xi′)σixi-xi′,θ(r)=-4r2h2,wherein J represents a point set of the point cloud sampling data, q represents the sampling points in the point set J, I represents a neighborhood point set of the sampling points q, x represents the neighborhood points in the neighborhood point set I, R is a regular term, γ is a weighting coefficient, h is a neighborhood radius of the neighborhood point set I, and σ is a distribution coefficient; and connecting the skeleton branches and obtaining a point cloud skeleton.