Vision-Based Horizontal Structure Tracking and Inspection via Unmanned Aerial Vehicle

The involvement of unmanned aerial vehicles in the civilian domains has made the task quite speedy, cost-effective, and risk-free. This paper provides a general approach to how horizontal civil structures such as bridges can be tracked and monitored using unmanned aerial vehicles. The work is divided into two sections. The first one is the identification of horizontal structures, followed by structure tracking. Images taken from the mounted camera on the UAV are first processed to extract features with the help of computer vision. Then, from the features, the important parameters are taken out to design a controller that handles the tracking process. Also, during this tracking process, the whole structure area is scanned. This work contains simulation results of vision-based structure tracking done on a gazebo simulator.