Continuous Structural Displacement Monitoring Using Accelerometer, Vision, and Infrared (IR) Cameras

Area of Science:

• Structural Engineering
• Computer Vision
• Sensor Technology

Background:

• Vision-based structural displacement measurement is limited by illumination changes and nighttime operation.
• Existing non-contact methods struggle with short-term measurements and environmental variability.

Purpose of the Study:

• To develop a continuous structural displacement estimation technique for both day and night operation.
• To overcome the limitations of vision-based methods using combined sensor data.
• To enable robust and long-term structural health monitoring.

Main Methods:

• Combined accelerometer, vision camera, and infrared (IR) camera measurements at the structure's displacement point.
• Automatic optimization of IR camera temperature range for feature matching.
• Adaptive reference frame updating for robust illumination-displacement estimation.

Main Results:

• Achieved continuous displacement estimation under varying conditions (day/night).
• Demonstrated high accuracy with root-mean-square error < 2 mm in lab tests.
• Validated field applicability on a pedestrian bridge.

Conclusions:

• The proposed technique enables continuous, long-term structural displacement monitoring with enhanced robustness.
• On-site sensor installation eliminates the need for fixed locations, making it attractive for practical applications.
• Future work could extend to multi-point displacement estimation by off-site camera installations.