Real time measurement of the dynamic displacement field of a large-scale arch-truss bridge by remote sensing technology

This paper describes the development of a computer vision-based real time displacement measurement system and demonstrated its performance on a large-scale wood truss bridge model. Digital images were captured with a consumer grade video camera. Three common types of computer vision algorithms are c...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE SENSORS 2016 : Orlando, Florida, USA, October 30-November 2, 2016 : 2016 proceedings papers S. 1 - 3
Hauptverfasser: Yang Yang, Xiong Bill Yu
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.10.2016
Schlagworte:
Algorithm design and analysis
Bridges
Computer vision
Displacement measurement
Fasteners
Integrated circuits
inverse compositional algorithm
Lucas-Kanade algorithm
remote sensing technology
Sensors
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:This paper describes the development of a computer vision-based real time displacement measurement system and demonstrated its performance on a large-scale wood truss bridge model. Digital images were captured with a consumer grade video camera. Three common types of computer vision algorithms are compared, including the Lucas-Kanade (LK) template tracking algorithm, inverse compositional (IC) algorithm (an extension of LK algorithm), and Digital Image Correlation (DIC). Application to the model bridge subjected to loading process indicates that the IC algorithm achieves real time displacement measurement. The performance in displacement from computer vision analyses matches the data collected by the conventional displacement sensors, with an average precision of within 1 mm at a distance of 5m away from the structure. The processing speed of the IC algorithm is over 300 faster than the conventional LK algorithm and around 140 times faster than Digital Image Correlation (DIC).
DOI:10.1109/ICSENS.2016.7808732