Multi-Target Computer-Vision Monitoring of Multi-Story Buildings During Shake Table Tests

Computer-vision methods have attracted much attention in recent years thanks to the possibility of measuring displacement time-histories in an unprecedented number of points using a single sensor not requiring stationary points close to the structure. The hardware has accessible costs and is simple to set up and operate with basic video photography knowledge and programming skills. Past experimental studies showed the potentiality of computer-vision algorithms to provide accurate results that can be equivalent to those obtained from other contact sensors as accelerometers and displacement transducers. This is made possible thanks to the development of sophisticated and computationally efficient computer-vision algorithms able to achieve resolutions as small as a fraction of a single pixel in the captured video footage, while being able to track also large displacements. The objective of this article is to present a vision-based methodology for contactless acquisition of absolute and relative horizontal displacement time-histories of multi-story buildings during shake table tests, using a couple of video cameras, one inside and the other outside of the monitored building. The combination of the two cameras provides redundancy of the measurements which can be used for the compensation of the inevitable disturbances in the internal video camera. The advantages of the proposed methodology are discussed based on the results of its first application that took place during the shake table testing of a full-scale six-story mass timber structure tested on the world’s largest 6-DOF outdoor shake table at the University of California San Diego.