Remote Inspection and Monitoring of Civil Engineering Structures Based on Unmanned Aerial Vehicles

Abstract

This chapter describes recent contributes for the remote inspection and monitoring of civil engineering structures with the support of Unmanned Aerial Vehicles (UAVs). Regarding the remote inspection, dedicated computer-vision methodologies are presented for the automatic detection of cracks and exposed steel rebars on reinforced concrete (RC) structures. Two distinct image processing techniques are used: one based on heuristic features and other based on deep learning. In case of deep learning the stages related to the image database, as well as the training and validation are detailed. Both techniques are successfully applied to the case study of a building facade and a large-scale silo structure. In relation to the monitoring, an innovative methodology for estimating in-plane structural displacements based on video systems and UAVs is presented. As the structure and the UAV are both in motion, the estimation of the absolute structural displacements involves, first, the evaluation of the relative displacements of the UAV-structure based on target tracking, and second, the UAV motion subtraction based on the data derived from an embedded Inertial Measuring Unit (IMU). The validation of the proposed methodology is based on an exploratory test performed in the field, which conducted to very satisfactory and promising results.