This essay analyzes the defects of target recognition in the field of unmanned reconnaissance, and gives an optimization strategy based on YOLOv5 algorithm. The feasibility of this system is verified by simulation experiment and real machine test, which has certain help to the battlefield unmanned reconnaissance. Then, the simulation experiment of unmanned reconnaissance system target detection and recognition was carried out, which mainly collected and established target data sets of armored vehicles, helicopters and people, and completed target recognition model training. After that, the parameters of the training results were analyzed, and the weights obtained from the training were tested. The algorithm was optimized and improved according to the characteristics of the uav targets in the ultra-low altitude perspective, and the model training and testing were carried out for many times, which improved the recognition rate of small and medium targets in the actual scene.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Design of high automatic target recognition unmanned reconnaissance system based on YOLOv5


    Beteiligte:
    Ji, Shuangxing (Autor:in) / Tang, Hua (Autor:in) / Ming, Yue (Autor:in) / Zhao, Chen (Autor:in)


    Erscheinungsdatum :

    2022-10-12


    Format / Umfang :

    1194247 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Subsonic unmanned air reconnaissance system design

    SWIFT, GERALD / SHEPARD, CHARLES / SEBEK, KEN | AIAA | 1990


    UNMANNED-FLIGHT RECONNAISSANCE SYSTEM

    CHAE HEE SEO / SONG HO NAM / PARK JOON HO | Europäisches Patentamt | 2016

    Freier Zugriff

    - Unmanned-flight reconnaissance system

    Europäisches Patentamt | 2019

    Freier Zugriff

    SE-yolov5 inland waterway vessel target recognition based on CGAN image enhancement

    Zhou, Wenhao / Ma, Yupeng / Xiao, Shengkui et al. | IEEE | 2023