In this paper we propose a robotic system for flower removal and pollination with the deep learning-based 3D perception module. The robotic system consists of a collaborative robot equipped with an RGB-D camera used in the detection pipeline. A semi-automated 3D based image annotation method (A3IA) is developed for the purpose of automatically generating the training data. The efficiency of the detection model trained on automatically annotated data is compared to that of the model trained on synthetic data. Experimental validation of the proposed methods is conducted and robustness and precision of both detection and positioning are reported.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Robotic Strawberry Flower Treatment Based on Deep-Learning Vision


    Weitere Titelangaben:

    Springer Proceedings in Advanced Robotics


    Beteiligte:
    Borja, Pablo (Herausgeber:in) / Della Santina, Cosimo (Herausgeber:in) / Peternel, Luka (Herausgeber:in) / Torta, Elena (Herausgeber:in) / Vuletić, Jelena (Autor:in) / Polić, Marsela (Autor:in) / Orsag, Matko (Autor:in)

    Kongress:

    International Workshop on Human-Friendly Robotics ; 2022 ; Delft, The Netherlands September 22, 2022 - September 23, 2022



    Erscheinungsdatum :

    02.01.2023


    Format / Umfang :

    16 pages





    Medientyp :

    Aufsatz/Kapitel (Buch)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch





    Design and Analysis of Strawberry-Picking Industrial Robotic Arm

    Amir, Adnan / Verma, Akshat / Goswami, Abhinav et al. | IEEE | 2022


    Vision-based robotic convoy driving

    Schneiderman, H. / Nashman, M. / Wavering, A. J. et al. | British Library Online Contents | 1995


    Robotic Vision System

    N. W. Witriol / W. H. Brumage | NTIS | 1986


    Robotic control using deep learning

    CHRISTOPHER JASON PAXTON / KEI KASE / DIETER FOX | Europäisches Patentamt | 2021

    Freier Zugriff