Noncontact deflection measurement for bridge through a multi‐UAVs system
Sheng Zhuge,Xiangpeng Xu,Lijun Zhong,Shuwei Gan,Bin Lin,Xia Yang,Xiaohu Zhang
DOI: https://doi.org/10.1111/mice.12771
2021-09-30
Abstract:Deflection measurement is the research focus of health monitoring for bridges during the operation period. This study develops a contactless measurement technique to monitor the bridge deflection, leveraging visual information from a team of unmanned aerial vehicles (UAVs). On the basis of the collinearity of the laser spots projected on the plane by the coplanar laser indicator, we can eliminate the motion of UAV, and calculate the vertical displacement of the position to be measured relative to the bridge pier. In the proposed method, the center of the laser spot is extracted through a method based on deep learning, and an algorithm based on scale-invariant features registration was developed to track the feature points of the bridge in the image sequence. According to the algorithm, we demonstrate the accuracy and feasibility of our approach through simulation and simulated bridge experiments. The result shows that the root mean squared error (RMSE) of measurement through our technique is less than 0.5 mm in the laboratory conditions. In addition, the limits and scalability of the presented method have been explored through a field experiment.
transportation science & technology,construction & building technology,computer science, interdisciplinary applications,engineering, civil