ePrints.FRI - University of Ljubljana, Faculty of Computer and Information Science

Computer-vision-based autonomous control for quadrocopter

Alan Lukežič (2012) Computer-vision-based autonomous control for quadrocopter. EngD thesis.

Download (6Mb)


    The main goal of the thesis is presenting the implementation of the mobile platform Parrot AR.Drone for object tracking. Parrot is a quadrocopter – an aerial vehicle similar to a helicopter, but with four propellers. As there is a camera attached to it and there is the possibility of wireless connection via laptop, the system belongs to the computer vision and robotics field. We created a system which is capable of autonomous tracking a manually selected object. For tracking we used two existent trackers and analyzed their suitability in real system. We integrated the trackers into the mobile platform and implemented suitable methods for navigating the quadrocopter in space. Using only the object's location in the image, the quadrocopter is able to continually estimate it's position in space. Using this piece of information our method regulate the flight route of the quadrocopter, so that it keeps the constant position to the object. We assessed the system with the help of empiric experiments on real examples of object tracking and measurements in controlled conditions. At the end we suggested improvements which should make the system more robust.

    Item Type: Thesis (EngD thesis)
    Keywords: visual object tracking, quadrocopter, Parrot AR.Drone, computer vision, robotics
    Number of Pages: 57
    Language of Content: Slovenian
    Mentor / Comentors:
    Name and SurnameIDFunction
    doc. dr. Matej Kristan4053Mentor
    Link to COBISS: http://www.cobiss.si/scripts/cobiss?command=search&base=50070&select=(ID=00009486932)
    Institution: University of Ljubljana
    Department: Faculty of Computer and Information Science
    Item ID: 1799
    Date Deposited: 12 Sep 2012 11:49
    Last Modified: 05 Nov 2012 13:16
    URI: http://eprints.fri.uni-lj.si/id/eprint/1799

    Actions (login required)

    View Item