Integrated Video Analysis Framework for Vision-Based Comparison Study on Structural Displacement and Tilt Measurements

AbstractWith the advancement and wide availability of digital video cameras, compounded with the development and enhancement of computer vision methods, vision-based sensing using a video camera as a sensor is promising to become a viable alternative and complementary approach to conventional structural health monitoring (SHM). Among many video analytics methods, no single method is effective for various practical SHM applications using a video camera as a sensor. Not only is it vitally important to understand the effectiveness and shortcomings of the methods, but also there is a need for developing an integrated framework and software tool to enable structural engineers to apply the suitable method for the SHM case at hand. To bridge the gap between the research and practical vision-based SHM applications, this paper presents a unified video analysis framework that integrates both the phase-based algorithm and template matching methods for processing SHM videos recorded using either industrial or consumer types of cameras. The methods have been implemented and integrated into a versatile software tool for extracting structural responses such as displacements, velocities, and accelerations at the points of interest on a structure. The integrated tool enables us to conduct a comparison study of displacement and tilt measurements for a reinforced concrete wall tested in a large structural laboratory and a highway bridge test in the field. The results obtained for the tests have been compared with measurements by conventional sensors. The comparison reveals the advantages and limitations of the phase-based algorithm and the template matching methods. The integrated framework, together with the lessons learned from the comprehensive comparison study, facilitates effective SHM applications using a video camera as a sensor in practice.