The aim of this Thesis is to research the possibility of docking an airborne UAV using a robotic
armand computer vision. In the scope of the SHERPA project on robot collaboration in Alpine
rescue scenario‘s a robust and autonomous way of retrieving small scale UAV‘s is required. In a
scenario where landing the UAV on a mobile ground station is not possible and landing on the
ground could be harmful to the drone, docking the UAV while it is airborne using a robotic arm
would be the best solution.
In this work a system is presented to accurately follow and dock an UAV near a ground vehicle
using an external monocular camera mounted on a robotic armand internal UAV sensors. The
system combines the strengths of visual pose estimation and IMU motion tracking while minimizing
the effect of their weaknesses. The result is a system with an accurate pose estimate
and which is also robust through temporary occlusion or motion blur. A functional proof of
principle system was implemented using Robotic Operating System and a KUKA LWR4+ robot
armto show the advantages provided by the system.
[1]
Bernhard P. Wrobel,et al.
Multiple View Geometry in Computer Vision
,
2001
.
[2]
Axel Pinz,et al.
Robust Pose Estimation from a Planar Target
,
2006,
IEEE Transactions on Pattern Analysis and Machine Intelligence.
[3]
Martin Buss,et al.
Visual tracking and control of a quadcopter using a stereo camera system and inertial sensors
,
2009,
2009 International Conference on Mechatronics and Automation.
[4]
Md. Arafatur Rahman,et al.
SHERPA: An Air-Ground Wireless Network for Communicating Human and Robots to Improve the Rescuing Activities in Alpine Environments
,
2014,
ADHOC-NOW Workshops.
[5]
Greg Welch,et al.
Welch & Bishop , An Introduction to the Kalman Filter 2 1 The Discrete Kalman Filter In 1960
,
1994
.
[6]
Davide Scaramuzza,et al.
A monocular pose estimation system based on infrared LEDs
,
2014,
2014 IEEE International Conference on Robotics and Automation (ICRA).