Aircraft Point Cloud Registration via Spherical Harmonic Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing aircraft assembly measurement technologies face challenges in efficiently registering multi-view point cloud data, leading to slowed registration speeds and potential precision issues due to error propagation and local optimization traps.
Innovation Solution
A method for multi-view point cloud registration based on spherical harmonic features (SHF) is proposed, involving data preprocessing, local spherical projection, calculation of SHF, constraint correspondence establishment, filtering, and optimization graph-based transformation to achieve accurate registration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pairwise registration is performed sequentially for multiple point clouds, then registration can be completed step by step, but registration error accumulates and precision deteriorates
Solution Approach 1:
The patent performs preliminary coarse registration to establish initial correspondence relationships between all point clouds before detailed optimization. This preliminary action provides a global framework that prevents error accumulation in sequential pairwise registration, as all point clouds are registered against a common reference frame established in advance.
Solution Approach 2:
The patent merges multiple point clouds into a unified coordinate system by establishing correspondence relationships between all overlapping point clouds simultaneously rather than sequentially. This combining approach allows global error optimization and prevents the accumulation of individual registration errors that occur in sequential processing.
2Measurement precision
If conventional point cloud registration is performed on large volume data, then complete coverage is achieved, but registration speed slows down
Solution Approach 1:
The patent segments the large volume point cloud data by establishing local correspondence relationships in overlapping regions between adjacent views. Instead of processing all points globally, the method divides the registration task into manageable local segments that can be processed efficiently and then integrated into a complete registration result.
Solution Approach 2:
The patent employs feature-point-based correspondence establishment that automatically identifies and matches key features without requiring manual intervention or complex global optimization. The method uses intrinsic geometric features of the point clouds themselves to drive the registration process, eliminating the need for external control points or markers.
3Productivity
If feature-based registration is used to speed up processing, then registration speed improves, but local optimization traps cause registration failure
Solution Approach 1:
The patent introduces an intermediary optimization graph that mediates between local feature matches and global registration consistency. The graph structure allows local feature-based correspondence to be established quickly while the graph optimization ensures global consistency, preventing the algorithm from getting trapped in local optima that would cause registration failure.
Data Source
AI summary
The present disclosure discloses a method for multi-view point cloud registration for a whole aircraft based on a spherical harmonic feature (SHF). Gaussian sphere projection is mainly performed on a local point cloud. An SHF of each point is obtained by using a spherical harmonic transform technology. A constraint correspondence between feature points is found based on an SHF of each point in the point cloud. Multi-view point cloud data registration is performed according to an optimization graph method under a constraint of the SHF. A key of multi-view point cloud registration is to search for feature constraint relationships between different observation stations and perform non-linear solution based on these relationships, so as to obtain a pose parameter of a point cloud in each viewing angle.