A Target-Less Vision-Based Displacement Sensor Based on Image Convex Hull Optimization for Measuring the Dynamic...

Area of Science:

• Structural Health Monitoring
• Computer Vision
• Mechanical Engineering

Background:

• Existing vision-based displacement sensors (VDSs) rely on identifiable targets with markers.
• Extracting displacement data without markers presents a significant challenge in structural monitoring.

Purpose of the Study:

• To propose and validate a target-less vision-based displacement sensor (TVDS).
• To demonstrate the TVDS's capability to extract displacement data and structural dynamic characteristics without markers.

Main Methods:

• Image convex hull optimization to identify and track feature points without targets.
• Calculation of a scaling factor map for converting pixel to physical coordinates.
• Validation through circular marker diameter estimation, white-noise excitation tests, and comparison with laser displacement sensors (LDS).

Main Results:

• The TVDS accurately extracts displacement data without requiring artificial or natural markers.
• Reliability of TVDS displacement data was confirmed by comparison with LDS measurements.
• Extracted dynamic characteristics (natural frequency, mode shape) closely matched numerical analysis results.

Conclusions:

• The proposed TVDS offers a reliable and accurate method for measuring structural displacement and dynamic characteristics.
• The target-less approach allows for flexible data extraction from any part of a structure.
• TVDS presents a significant advancement for structural health monitoring applications.