Traditional artifact detection needs to manually select and extract suitable image features, but some artifacts have weak surface textures, which may cause problems such as difficulty in feature extraction or poor robustness. In order to apply visual servoing technology to the detection and positioning of low-texture artifacts, this paper proposes a target detection platform based on YOLOv4-Tiny. Using four types of low-texture water pipe connectors with different shapes as image samples, a data set was made, which not only realized the accurate identification of low-texture artifacts, but also reduced the actual hardware requirements. For the positioning of the target artifact, the relative pose estimation method based on the target detection frame is proposed for further improving the visual servo system. The experimental results show that the system solves the problem of difficulty in extracting visual features of low-texture artifacts, and the positioning error can be controlled within ±1 mm or ±1 degree in practical application.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Low Texture Artifacts Recognition and Visual Servo Position Based on YOLOv4-Tiny


    Beteiligte:
    Lin, Jun (Autor:in) / Luo, Jiaguo (Autor:in) / Liu, Xianghui (Autor:in) / Zhong, Xungao (Autor:in) / Yang, Guizhi (Autor:in) / Zhao, Jing (Autor:in)


    Erscheinungsdatum :

    2022-10-12


    Format / Umfang :

    1616617 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Fast Detection of Railway Fastener Using a New Lightweight Network Op-YOLOv4-Tiny

    Liu, Jianwei / Qiu, Yuan / Ni, Xuefeng et al. | IEEE | 2024




    A Modified YOLOv4 Deep Learning Network for Vision-Based UAV Recognition

    Farzaneh Dadrass Javan / Farhad Samadzadegan / Mehrnaz Gholamshahi et al. | DOAJ | 2022

    Freier Zugriff

    Pedestrian Detection Algorithm Based on Improved YOLOv4

    Zheng, Ziheng / Ni, Chenhui / Zeng, Guolei | IEEE | 2023