Visual Servoing of a Quadrotor UAV for Autonomous Power Lines Inspection

The objective of this work is to achieve real time tracking of power conductors using a Quadrotor Unmanned Air Vehicle (UAV). Two solutions are proposed. For the first solution, the control is directly solved in the 2 D image space. The link between image features and the kinematics of the vehicle is achieved using an Image Based Visual Servoing (IBVS) formulation, combined with a Linear Quadratic Servo (LQ-Servo) approach. In the second solution, a partial pose of the vehicle with respect to (wrt) the lines is estimated. Then, a position controller designed in the Cartesian space is used to drive the vehicle to the desired configuration. Experimental results are provided to validate the two approaches.