A Vision-Based Sensor for Noncontact Structural Displacement Measurement

Area of Science:

• Structural Engineering
• Computer Vision
• Measurement Science

Background:

• Conventional displacement sensors face practical limitations in structural monitoring.
• Remote and non-contact measurement methods are needed for enhanced structural health assessment.

Purpose of the Study:

• To develop and validate a vision sensor system for real-time, remote measurement of structural displacements.
• To assess the accuracy and feasibility of the vision sensor in laboratory and field conditions.

Main Methods:

• Development of a vision sensor system employing an advanced template matching algorithm (upsampled cross-correlation).
• Real-time displacement extraction from video images with adjustable subpixel resolution.
• Laboratory validation using a shaking table test on a frame structure.
• Field tests on a railway bridge and a pedestrian bridge.

Main Results:

• The vision sensor system demonstrated satisfactory agreement with high-performance laser displacement sensors in laboratory tests.
• Field tests confirmed the accuracy of the vision sensor in both time and frequency domains under realistic conditions.
• The system successfully tracked displacements using both artificial and natural targets on structural surfaces.

Conclusions:

• The developed non-contact vision sensor system offers a cost-effective, user-friendly, and flexible solution for structural displacement monitoring.
• This technology provides accurate remote measurement capabilities, enhancing structural health assessment and safety.
• The system's ability to extract displacement data from any point using a single camera represents a significant advancement.