Autonomous navigation for large Unmanned Aerial Vehicles (UAVs) is fairly straight-forward, as expensive sensors and monitoring devices can be employed. In contrast, obstacle avoidance remains a challenging task for Micro Aerial Vehicles (MAVs) which operate at low altitude in cluttered environments. Unlike large vehicles, MAVs can only carry very light sensors, such as cameras, making autonomous navigation through obstacles much more challenging. In this paper, we describe a system that navigates a small quadrotor helicopter autonomously at low altitude through natural forest environments. Using only a single cheap camera to perceive the environment, we are able to maintain a constant velocity of up to 1.5m/s. Given a small set of human pilot demonstrations, we use recent state-of-the-art imitation learning techniques to train a controller that can avoid trees by adapting the MAVs heading. We demonstrate the performance of our system in a more controlled environment indoors, and in real natural forest environments outdoors.


    Access

    Download


    Export, share and cite



    Title :

    Learning Monocular Reactive UAV Control in Cluttered Natural Environments



    Publication date :

    2012


    Remarks:

    8 pages, 10 figures


    Type of media :

    Preprint


    Type of material :

    Electronic Resource


    Language :

    English




    Vision and Learning for Deliberative Monocular Cluttered Flight

    Dey, Debadeepta / Shankar, Kumar Shaurya / Zeng, Sam et al. | ArXiv | 2014

    Free access

    IMITATION LEARNING FOR UAS NAVIGATION IN CLUTTERED ENVIRONMENTS

    Harris, Caleb M. / Choi, Youngjun / Mavris, Dimitri N. | TIBKAT | 2021


    Imitation Learning for UAS Navigation in Cluttered Environments

    Harris, Caleb M. / Choi, Youngjun / Mavris, Dimitri N. | AIAA | 2021


    Aerial Locomotion in Cluttered Environments

    Floreano, Dario / Zufferey, Jean-Christophe / Klaptocz, Adam et al. | British Library Conference Proceedings | 2011


    Robot navigation in cluttered environments

    Regier, Peter | TIBKAT | 2022

    Free access