Change Analysis in Structural Laser Scanning Point Clouds: The Baseline Method

Area of Science:

• Geomatics Engineering
• Structural Health Monitoring
• Geospatial Data Analysis

Background:

• Change detection in point clouds typically requires registration, a process prone to errors and demanding stable features.
• Existing methods for analyzing structural changes from 3D scans face limitations due to registration inaccuracies.
• A need exists for a registration-free method to reliably detect subtle changes in scanned environments.

Purpose of the Study:

• To develop and validate a novel change detection technique for point clouds that bypasses the traditional registration step.
• To assess the efficacy of a baseline-based approach for identifying structural deformations, specifically after seismic events.
• To compare the proposed method's performance against conventional cloud-to-cloud analysis for structural integrity assessment.

Main Methods:

• A registration-free change detection method is proposed, utilizing 'baselines' that connect feature points within a single scan.
• Baselines connecting corresponding feature points (targets or virtual points) across two scans are compared to detect changes.
• The method was applied to point cloud data of a masonry laboratory building before and after seismic testing, using automatically extracted brick centers as virtual points.

Main Results:

• The baseline analysis successfully detected structural damages in the masonry laboratory building, with deformations measured in centimeters.
• The proposed method demonstrated its potential for accurate structural analysis by identifying changes without relying on registration.
• Comparison with traditional cloud-to-cloud change detection highlighted the advantages of the baseline-based approach.

Conclusions:

• The developed baseline-based method offers a robust and accurate alternative for detecting changes in point clouds, particularly for structural monitoring.
• This registration-free technique mitigates errors associated with alignment and expands the applicability of 3D scan data analysis.
• The method shows significant promise for assessing structural health and damage in civil engineering applications.