Two flexible vision-based methods for remote deflection monitoring of a long-span bridge

Abstract Two different vision-based methods for remote and multipoint deflection measurement of large-scale bridges are developed. In Method 1, the deflection is obtained by detecting the motions of bridge lights through a camera under the bridge. A toolless object distance measurement method is proposed for magnification factor calibration. For bridges spanning the water or a deep canyon, Method 2 designs to monitor deflection using a camera placed on the bridge deck. A monitoring model that only considers the camera attitudes related to the vertical displacement measurement is adopted for disturbance correction. The efficacy and flexibility of two methods for remote bridge displacement monitoring were demonstrated through indoor verification tests and a field experiment on a long-span cable-stayed bridge. Both proposed methods have great potential in the safety evaluation of various bridges and other large-scale engineering structures.