An Advanced Edge-Detection Method for Noncontact Structural Displacement Monitoring

Area of Science:

• Structural Engineering
• Computer Vision
• Non-contact Measurement

Background:

• Traditional methods for structural monitoring often require physical contact or marked targets.
• Integer pixel edge detection methods can lack the precision needed for subtle structural changes.
• Developing cost-effective and accurate non-contact monitoring solutions is crucial for infrastructure health assessment.

Discussion:

• This paper proposes a novel non-contact vision sensor system leveraging advanced Zernike subpixel edge detection.
• The system effectively detects structural features and displacements without requiring manually placed targets.
• Integration of subpixel techniques enhances accuracy and cost-effectiveness compared to traditional pixel-based methods.

Key Insights:

• The vision sensor system demonstrated satisfactory agreement with established Multipurpose Testing System (MTS) measurements.
• Field tests on a street sign confirmed the system's ability to monitor multiple points simultaneously without targets.
• The system instantaneously provides natural frequencies and mode shapes, aiding in damage localization.

Outlook:

• Further validation of the vision sensor system across diverse structural types and environmental conditions is recommended.
• Potential integration with real-time structural health monitoring platforms could enhance infrastructure management.
• Exploration of advanced algorithms for automated damage detection and classification based on modal data is a promising future direction.