Spiral Trajectories for Building Inspection with Quadrotors

Inspection of large buildings is an important task since it can prevent material and human losses. A cheap and fast way to do the inspection is with sensors mounted on quadrotor vehicles. The challenge here is to compute a trajectory so that the building is completely observed while this same trajectory can be followed by the quadrotor in a smooth way. To address the problem, we propose a method that receives a 2.5D model of the target building and computes a smooth trajectory that can be followed by the quadrotor controller. The computed trajectory is a Fourier series that matches the desired behaviour. The advantage of the proposed method is that we can select the Fourier components in order to decrease the high frequency movement. The method has been tested in simulation and we have compared it against polynomial trajectories. Our results show that the method is efficient and can be applied to different building shapes.