Point Cloud Scan Merging via Segmentation and Feature Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for aligning multiple laser scan point cloud data sets are inefficient, as each scan has a unique vantage point and does not cover the entire area, making it difficult to determine accurate overlap and alignment for comprehensive measurement.
Innovation Solution
The system imports and segments scans, filters geometric attributes, and uses these segments for registration by matching and aligning them in a geometric manner, employing techniques like box-filtering, least-squares estimation, and iterative closest point (ICP) methods to refine transformations and score matches for accurate alignment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple laser scan range images are used to capture a desired area, then the coverage area is improved, but the alignment and registration complexity increases
Solution Approach 1:
The patent segments the point cloud data into multiple regions or features, allowing each scan to be processed and aligned independently through feature matching, thereby managing the complexity of aligning multiple scans while maintaining comprehensive area coverage
Solution Approach 2:
The patent introduces coordinate transformation and registration algorithms as intermediary processes that automatically align multiple scans by matching common features across different scans, reducing the manual alignment complexity while preserving the expanded coverage area
2Measurement precision
If multiple scans are aligned to create a single comprehensive image, then the measurement accuracy is improved, but the computational load increases
Solution Approach 1:
The patent divides the point cloud data into segmented features or regions, allowing the registration process to work with smaller, manageable segments rather than processing all points from every scan simultaneously, thus reducing computational load while maintaining measurement accuracy through precise feature matching
Solution Approach 2:
The patent performs preliminary processing of scan data including filtering, downsampling, and feature extraction before the main registration process, which reduces the computational burden of the alignment operation while preserving the accuracy needed for comprehensive measurement
Data Source
AI summary
A method, system, apparatus, article of manufacture, and computer-readable storage medium provide the ability to merge multiple point cloud scans. A first raw scan file and a second raw scan file (each including multiple points) are imported. The scan files are segmented by extracting segments based on geometry in the scene. The segments are filtered. A set of candidate matching feature pairs are acquired by registering features from one scan to features from another scan. The two raw scan files are merged based on the candidate matching feature pairs.