3D Shape Matching Using Spatial Geometry Histograms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 3D local feature descriptors are limited in their ability to provide detailed information, are often scene-specific, and are sensitive to noise and grid resolution, leading to reduced accuracy in describing 3D local surfaces.
Innovation Solution
A 3D shape matching method based on 3D local feature description using Spatial Geometry Histograms (SGHs), which involves establishing a local reference frame for a spherical neighborhood around a feature point, dividing it into radial, azimuth, elevation, and deviation partitions, and generating corresponding histograms to characterize 3D local surface information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If local feature descriptors based on LRA (Local Reference Axis) are used, then the descriptor structure is simple, but the descriptiveness is insufficient due to limited radial and elevation direction information
Solution Approach 1:
The patent transitions from LRA (single axis, 2D histogram) to LRF (three orthogonal axes, 3D histogram), adding a dimensional aspect to the feature descriptor. This enables encoding of spatial distribution information in three directions (radial, azimuth, elevation), significantly improving descriptiveness while maintaining manageable complexity through systematic histogram binning
2Ease of manufacture
If existing local feature descriptors are used, then the algorithm is simple to implement, but the accuracy is reduced due to sensitivity to noise and grid resolution
Solution Approach 1:
The patent employs multiple parameter optimization strategies: (1) adaptive selection of support radius R based on local curvature, (2) dynamic adjustment of histogram binning resolution, (3) normalization of histogram values to reduce sensitivity to grid resolution variations. These parameter changes maintain implementation simplicity while significantly improving robustness to noise and resolution variations
3Productivity
If existing local feature descriptors are used, then the computational process is fast, but the descriptors are limited to specific scenes and lack generalizability
Solution Approach 1:
The patent creates a universal descriptor framework based on LRF that can adapt to different scene types (indoor, outdoor, structured, unstructured environments). The three-axis histogram approach with configurable binning and support radius makes the descriptor scene-agnostic while maintaining computational efficiency through standardized histogram computation and comparison operations
Data Source
AI summary
A 3D shape matching method and a 3D shape matching device based on 3D local feature description using SGHs are provided. In the method, the spherical neighborhood of the feature point is not only divided based on space but also divided based on geometry, the spherical neighborhood of the feature point is not only divided based on the radial direction and the azimuth respectively but also divided based on the elevation, and the spherical neighborhood of the feature point is not only divided based on the deviation angle deviating from the z axis but also divided based on the deviation angle deviating from the x axis. When the deviation angle deviating from the z axis of the spherical neighborhood is divided, the deviation angle is divided more densely where it is closer to the positive direction of the z axis.