Non-Contact Vision Based Structural Vibration Measurements: A Comparative Analysis on Tracker Algorithms for Structural Health Monitoring

Veen, Huib van der (2024) Non-Contact Vision Based Structural Vibration Measurements: A Comparative Analysis on Tracker Algorithms for Structural Health Monitoring. Integration Project, Industrial Engineering and Management.

Preview	Text FRIEMIPHuibvanderVeenS4017021.pdf Download (13MB) | Preview
	Text toestemming_ Huib Van der Veen _ opleiding_ Industrial Engineering and Management.pdf Restricted to Registered users only Download (135kB)

Abstract

This research evaluates the efficiency of various object-tracking algorithms compared to traditional accelerometer measurements in non-contact, vision-based structural vibration analysis. Specifically, the study focuses on validating the use of a smartphone for monitoring vibrations on a cantilever beam, assessing data quality in both time and frequency domains. The study explores the performance of different tracking algorithms, including the CSRT, MIL, KCF, and OF trackers, in capturing structural vibrations. Findings reveal that the OF tracker shows promise due to its high accuracy and sensitivity, as evidenced by its excellent signal-to-noise ratio and precise identification of primary frequency components. A real- world application on a lamppost further explores the potential of the OF tracker in practical scenarios. These findings highlight the potential of vision-based monitoring systems as valuable tools for structural health monitoring.

Item Type:	Thesis (Integration Project)
Supervisor name:	Cheng, L. and Kloosterman, H. and Xie, K.
Degree programme:	Industrial Engineering and Management
Thesis type:	Integration Project
Language:	English
Date Deposited:	10 Jul 2024 12:17
Last Modified:	10 Jul 2024 12:17
URI:	https://fse.studenttheses.ub.rug.nl/id/eprint/33185

Actions (login required)

View Item