Work

Camera-Based Dynamic Vibration Measurement Using CoTracker Model

Year: 2024

Type: conference-paper

Abstract: This paper investigates the CoTracker model, a sophisticated deep learning approach based on transformer architecture, for its application in dynamic vibration measurement using vision cameras. Compared to traditional accelerometers, which are costly and limited to point-based measurements, CoTracker promises a cost-effective, full-field analysis by tracking pixel movements in video sequences. We applied this model to analyze the vibrations of a cantilever beam captured through a high-speed camera, comparing its output against data from a standard reference accelerometer. Our findings suggest that CoTracker could revolutionize structural vibration monitoring due to its efficiency and broad applicability.

Source: International Conference on Industrial Artificial Intelligence (IAI)

DOI: 10.1109/iai63275.2024.10730170

Authors: Liangliang Cheng, Justin de Groot, Kun Xie, Yunpeng Zhu

Institutions: University of Groningen, Queen Mary University of London

Language: English

Topic: Structural Health Monitoring Techniques

Subfield: Civil and Structural Engineering

Field: Engineering

Domain: Physical Sciences

SDG: Industry, innovation and infrastructure

Open Access status: green

Funder: China Scholarship Council