A central issue for the autonomous navigation of mobile robots is to map unknown environments while simultaneously estimating its position within this map. This chicken-eggproblem is known as simultaneous localization and mapping (SLAM). Asctec’s quadrotor Pelican is a powerful and flexible research UAS (unmanned aircraft system) which enables the development of new real-time on-board algorithms for SLAM as well as autonomous navigation. The relative UAS pose estimation for SLAM, usually based on low-cost sensors like inertial measurement units (IMU) and barometers, is known to be affected by high drift rates. In order to significantly reduce these effects, we incorporate additional independent pose estimation techniques using exteroceptive sensors. In this article we present first pose estimation results using a stereo camera setup as well as a laser range finder, individually. Even though these methods fail in few certain configurations we demonstrate their effectiveness and value for the reduction of IMU drift rates and give an outlook for further works towards SLAM.


    Access

    Download


    Export, share and cite



    Title :

    TOWARDS STEREO VISION- AND LASER SCANNER-BASED UAS POSE ESTIMATION



    Publication date :

    2015-08-28


    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Classification :

    DDC:    000 / 510 / 629




    TOWARDS STEREO VISION- AND LASER SCANNER-BASED UAS POSE ESTIMATION

    Kersten, Jens / Rodehorst, Volker | BASE | 2015

    Free access

    STEREO VISION-BASED DRIVER HEAD POSE ESTIMATION

    Hoffken, M. / Tarayan, E. / Kressel, U. et al. | British Library Conference Proceedings | 2014



    An Efficient Approach to Onboard Stereo Vision System Pose Estimation

    Sappa, A.D. / Dornaika, F. / Ponsa, D. et al. | IEEE | 2008