Aiming at the landing problem of Unmanned Aerial Vehicle (UAV) without landmarks and global coordinates, this paper proposes an efficient vision-based UAV autonomous landing solution. Firstly, image information is obtained from the bottom visual sensor of the UAV, and the YOLOV5 object detection method is employed to determine the initial position of the dynamic apron. Subsequently, an improved dynamic apron tracking method based on Spatio-Temporal Transformer is designed. This method updates an initial template through target detection algorithm and utilizes multi-source perceptual data for confidence evaluation and dynamic template updates. It outputs the distribution estimation of the predicted bounding box corner coordinates, and achieves the high reliability continuous positional smoothing estimation and dynamic apron loss re-detection based on UAV vision. Finally, the UAV calculates the horizontal control quantity based on the horizontal relative position between the UAV and the apron obtained. The vertical control quantity is calculated based on the vertical relative position between the UAV and the apron. The vertical relative position is estimated using laser ranging and visual odometry techniques. Based on the results of horizontal tracking control and vertical regulation, velocity commands for the UAV in all three directions are generated, enabling the UAV to achieve autonomous landing on the dynamic apron. This paper achieves precise landing for dynamic apron in real-world environments. The experimental results demonstrate that this method avoids the limitations of labeled data and global information on autonomous landing of UAV. Even under GPS-denied conditions, the UAV can generate control commands based on visual data and achieve autonomous landing for randomly moving apron.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Vision-based Autonomous Landing of UAV on Dynamic Apron


    Contributors:
    Yue, Pengyu (author) / Xin, Jing (author) / Mao, Zhijie (author) / Lu, Yongchang (author) / Zhang, Youmin (author)


    Publication date :

    2023-10-13


    Size :

    1065969 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Unmanned aerial vehicle take-off and landing parking apron

    ZHU SHIBING / LIANG HUIHONG | European Patent Office | 2024

    Free access

    Autonomous landing of airplanes by dynamic machine vision

    Schell, F.-R. / Dickmanns, E. D. | British Library Online Contents | 1994


    Tunnel vision for a safer apron

    Online Contents | 2008


    Vision-Based Algorithm for Autonomous Aerial Landing

    Morando, A. E.S. / Santos, M. Ferreira / Castillo, P. et al. | IEEE | 2024


    COMPUTER-VISION-BASED AUTONOMOUS OR SUPERVISED-AUTONOMOUS LANDING OF AIRCRAFT

    ROBERT MERRILL CRAMBLITT | European Patent Office | 2020

    Free access