Characterizing Deformation of Buildings from Videos

Abstract

We have started to explore the feasibility of extracting useful data on the deformation of buildings and structures based on optical videos, (Taghavi Larigani & Heaton). In the beginning, we look at the characterizations and limitations of the hardware, which is composed of a high-quality digital camera, combined with its optical imaging system capturing a video-footage of the structure under test, and then introduce a straightforward targets-tracking algorithm that produces the time-series displacements of targets that we select on the video. We performed preliminary measurements consisting of testing our targets-tracking algorithm using high definition format videos displaying the structures that we wanted to test. The measurements pertain to a 1) finite-element software-generated video of JPL/NASA principal building, 2) YouTube-video of a seismic dynamic test of a building, 3) YouTube-video of the Millennium London Bridge “Wobbly Bridge”, 4) YouTube-video of a United Boeing 777, 4) YouTube-video of NASA space shuttle rockets during launch. So far, our tests are encouraging. If our approach proves viable, it can be transformative for the field of earthquake engineering and structural health monitoring. Hence, we consider the prospect of using our technique for surveying buildings and other civil structures in high seismic risk urban agglomerations. In parallel, the same technique could be applied for 1) real-time structural health monitoring of civil structures, 2) nuclear plants, 3) oil and gas infrastructures, 4) rail & road networks, 5) aircraft, 6) spacecraft, 7) etc., by simply analyzing the structure-facing camera recorded data.