Vision-Based Building Seismic Displacement Measurement by Stratification of Projective Rectification Using Lines

Area of Science:

• Structural Engineering
• Computer Vision
• Geophysics

Background:

• Accurate seismic displacement measurement is crucial for structural health monitoring.
• Existing vision-based methods often require fixed reference points or full camera calibration, limiting their practical application.

Purpose of the Study:

• To develop a flexible and accurate vision-based technique for seismic displacement measurement of building structures.
• To enable measurement using a single non-stationary camera with any perspective view, reducing prior information requirements.

Main Methods:

• Employs geometric constraints in the world coordinate system for projective rectification.
• Utilizes a stratified approach to determine projective transformation from line-based geometric relationships.
• Leverages plentiful line features in architectural environments for distortion removal.

Main Results:

• Demonstrates effective and robust performance through simulations and real-recorded data.
• Achieves accurate vision-based seismic displacement measurement without a priori camera parameters or fixed reference points.
• Successfully removes projective/affine distortion using line features.

Conclusions:

• The proposed method offers a practical advancement for vision-based structural measurement.
• It facilitates the transition of consumer-grade camera systems from laboratory settings to real-world structural health monitoring.
• Enhances the capability of non-stationary cameras for accurate seismic displacement analysis.