Plane-Fitting Point Cloud Registration for Noise-Robust Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing point cloud registration methods face high computational complexity and sensitivity to noise due to the need for point-by-point feature calculation and comparison, which is exacerbated by noise in the point cloud data.
Innovation Solution
A point cloud registration method based on plane fitting, where two non-parallel planes are identified in each point cloud, and a rotation matrix and translation vector are calculated to align the point clouds, reducing reliance on local point features and noise sensitivity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If point-by-point feature calculation and comparison is used for coarse registration, then registration accuracy can be achieved, but computational complexity increases significantly
Solution Approach 1:
The patent extracts plane information from the point cloud data, focusing on planar surfaces rather than individual point features. By identifying planes and using their geometric properties (normal vectors, equations) for registration, the method reduces computational complexity while maintaining accuracy on planar structures.
Solution Approach 2:
The patent changes the registration parameters from point-based features to plane-based parameters. Instead of calculating features for each point and comparing them, the method uses plane equations and normal vectors as the basis for transformation calculation, significantly reducing the computational burden.
2Measurement precision
If point features are calculated using neighborhood point cloud information, then feature representation is improved, but sensitivity to noise increases
Solution Approach 1:
The patent extracts plane information from the point cloud, which provides a more robust representation than individual point features. Planes are inherent geometric structures that can be identified even in the presence of noise, making the registration less sensitive to noisy points compared to point-based feature methods.
Solution Approach 2:
The patent introduces plane fitting as an intermediary step between the point cloud and the registration process. By first fitting planes to the point cloud data and then using these planes for registration, the method filters out the influence of noise while preserving the essential geometric information needed for accurate registration.
Data Source
AI summary
The present disclosure provides a point cloud registration method, an apparatus and electronic equipment based on plane fitting, after obtaining the source point cloud and the to-be-registered point cloud, two non-parallel source planes in the source point cloud and two non-parallel to-be-registered planes in the to-be-registered point cloud are determined, a rotation matrix is calculated based on the two source planes and the two to-be-registered planes, and a translation vector is obtained based on the source point cloud and the to-be-registered point cloud. Finally, the to-be-registered point cloud is rotated and translated according to the rotation matrix and the translation vector, to realize a registration between the to-be-registered point cloud and the source point cloud. The present solution realizes the registration based on a correspondence between planes, rather than based on local features of points, which can greatly reduce amount of calculation and sensitivity to noise.