Calibration and Set-up of Two Independent Cameras Mounted on Servos for UAVs Relative Pose Estimation

Unmanned Aerial Vehicles (UAVs) are agile platforms that are used in robotics for a wide range of tasks. However, due to their ability to perform fast flights, they are challenging to track. In literature there exists a broad number of tracking algorithms that allow to estimate the pose of an UAV by means of an external camera. Generally these algorithms make use of a fixed camera, and therefore they loose track of the UAV if it moves out of the field of view. A possible solution to this issue is to allow the presence of a moving camera. In this semester project, a system composed by a camera and a servo is considered. In particular, a toolbox for the calibration of the camera-servo set-up is proposed and analysed. First, the formalization of the problem is introduced. Then, the pipeline of the overall calibration is explained and evaluated. The proposed toolbox shows satisfactory performances, even when used in combination with algorithms for UAV pose estimation and tracking.