3D Scanning Calibration via Point Cloud Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current scanning systems face challenges in achieving high accuracy while maintaining ease of use, as they often require complex 3D references with specific features for calibration, which are costly and time-consuming to manufacture and configure.
Innovation Solution
A scanning system and method that uses a 3D-reference for calibration without specific feature requirements, employing a processor to generate a three-dimensional representation of a calibration object, register images, transform coordinates, and align point clouds to achieve high accuracy without the need for corner detection or nonlinear optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional calibration methods using 3D references with specific features (checker patterns, orthogonal planes) are employed, then measurement precision is improved, but device complexity and manufacturing cost increase
Solution Approach 1:
The patent extracts the essential calibration function from complex 3D references with specific features and implements it through simple point cloud registration. The method removes the need for predetermined features like checker patterns by directly registering scanned point clouds to reference models, keeping only the necessary geometric correspondence.
Solution Approach 2:
The patent uses digital copies of reference objects (CAD models, reference point clouds) instead of physical references with specific features. The virtual reference models are registered against scanned data through point cloud alignment, eliminating the need for physical checker patterns or specially manufactured calibration artifacts.
2Measurement precision
If complex 3D references with predetermined features are used for calibration, then measurement precision is improved, but ease of manufacture deteriorates
Solution Approach 1:
The patent replaces expensive, precisely manufactured calibration artifacts with simple, easily manufactured references. Any object with a known digital model can serve as a calibration reference, eliminating the need for costly fabrication of checker patterns or precision mechanical artifacts.
Solution Approach 2:
The calibration method becomes universal and applicable to any object with a known digital model. The same point cloud registration approach works for diverse reference objects (dental models, mechanical parts, artifacts) without requiring object-specific calibration procedures or specialized manufacturing.
3Measurement precision
If corner detection and nonlinear optimization are employed in calibration, then measurement precision is improved, but productivity deteriorates
Solution Approach 1:
The patent skips the time-consuming steps of corner detection and nonlinear optimization by directly performing point cloud registration. The method rushes through the calibration process using efficient point-to-plane or point-to-point registration algorithms that provide sufficient accuracy without the computational overhead of iterative optimization.
Solution Approach 2:
The patent replaces the mechanical/correspondence-based corner detection system with a direct point cloud registration system. Instead of detecting corners and performing geometric computations, the method uses spatial transformation of point clouds, substituting complex computational mechanics with simpler spatial alignment operations.
4Ease of operation
If calibration is performed in 2D-space using image coordinates, then ease of operation is improved, but measurement precision deteriorates
Solution Approach 1:
The patent transitions calibration from 2D image space to 3D point cloud space. By performing registration in three-dimensional space using spatial coordinates rather than image coordinates, the method achieves superior measurement precision while maintaining operational simplicity through automated point cloud processing.
Data Source
AI summary
Disclosed is a non-calibrated scanning system to be modified to a calibrated scanning system based on generating a three-dimensional representation of a three-dimensional calibration-object.