Camera motion-free computer vision-based displacement measurement

Abstract

Structural health monitoring (SHM) for bridges often requires displacement that can provide essential information on structural integrity. Nevertheless, displacement measurement on a full‐scale bridge is challenging due to inability to find the sensor installation point, insufficient resolution, or harsh field conditions. Recently, computer vision has been developed as a powerful tool for measuring displacement that offers easy-to‐find installation site with sub‐mm resolution. However, displacement measured by the computer vision is severely disturbed by camera motions especially when the camera is remotely installed. This study proposes a camera motion‐free computer vision‐based displacement measurement method by introducing an additional camera called a rear camera. The rear camera is designed to measure the 6 degrees‐of‐freedom (DOF) motions of the rear camera. Attaching the rear camera to the computer vision system, the 6‐DOF motion of the vision system can be indirectly estimated by the rear camera motions. Deriving the relationship between the motion‐induced errors with the rear camera motions, the motion‐free structural displacement is calculated. This article is organized by the introduction of the developed computer vision‐based displacement measurement method and a laboratory‐scale validation.