Building Datum Extraction from Laser Scanning Point Clouds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for extracting building datum information from point cloud data generated by laser scanners are labor-intensive and error-prone, failing to accurately represent the whole floor plan structure due to reliance on single horizontal slices and inability to filter out clutter, leading to incomplete and inaccurate BIMs.
Innovation Solution
A computer-implemented method that estimates principal axis directions, transforms point cloud data to make buildings stand upright, organizes data into a 3D voxel structure, and uses histogram analysis and plane-sweeping methods to extract level and orthogonal reference grid information, filtering out clutter and improving data accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single horizontal slice method is used to extract floor plan information, then the extraction process is simplified, but the completeness and accuracy of the floor plan structure is compromised
Solution Approach 1:
The patent divides the building point cloud data into multiple horizontal slices at different heights, then processes each slice separately to extract wall, floor, and ceiling information. This segmentation approach ensures comprehensive coverage of the entire building structure while maintaining manageable processing complexity for each individual slice.
Solution Approach 2:
The patent extends the extraction process from a single 2D horizontal slice to multiple 2D slices stacked in the vertical dimension, effectively creating a 3D representation through multiple 2D processing steps. This dimensional extension captures the complete building structure without overwhelming computational complexity.
2Productivity
If automated extraction methods are implemented, then productivity is improved, but accuracy and reliability of the extracted data deteriorates due to clutter and incomplete representation
Solution Approach 1:
The patent applies different processing strategies to different regions within each horizontal slice. Wall regions are identified and processed differently from floor and ceiling regions, and clutter detection is applied selectively to remove unwanted objects while preserving structural elements. This localized processing maintains high accuracy while enabling automated extraction.
Solution Approach 2:
The patent implements iterative refinement processes where extracted features are validated and refined through multiple passes. The system detects walls, floors, and ceilings in each slice, then uses this information to improve subsequent extractions, progressively enhancing accuracy while maintaining automated operation.
3Loss of information
If multiple scans are used to capture the complete point cloud, then coverage and completeness are improved, but the processing time and computational resources increase
Solution Approach 1:
The patent divides the large point cloud dataset from multiple scans into multiple manageable horizontal slices, processing each slice independently and in parallel where possible. This segmentation reduces the computational burden of processing the entire point cloud at once while maintaining complete building coverage through comprehensive slice coverage.
Solution Approach 2:
The patent processes each horizontal slice to the level of detail needed for that specific region, rather than applying exhaustive processing to the entire building at once. This partial action approach maintains necessary accuracy for structural elements while reducing overall processing time through selective detail levels in different regions.
Data Source
AI summary
A method, apparatus, system, and computer program product provide the ability to extract level information and reference grid information from point cloud data. Point cloud data is obtained and organized into a three-dimensional structure of voxels. Potential boundary points are filtered from the boundary cells. Level information is extracted from a Z-axis histogram of the voxels positioned along the Z-axis of the three-dimensional voxel structure and further refined. Reference grid information is extracted from an X-axis histogram of the voxels positioned along the X-axis of the three-dimensional voxel structure and a Y-axis histogram of the voxels positioned along the Y-axis of the three-dimensional voxel structure and further refined.