The role of agricultural development is very important in the economy of a country. However, the prevalence of numerous plant diseases significantly slows down crop growth and reduces its quality. Previously, plant diseases and quality of the plants were detected using digital image processing, recently deep learning break through produced better results compared with traditional approach. Due to the presence of low-contrast information in the input samples, the precise identification and classification of crop leaf diseases is a difficult and time-consuming process. The presence of noise and blurriness effects in the input images as well as changes in the size, location, and structure of the crop add to the difficulty of the classification process. A reliable drone-based deep learning approach is suggested in order to address the shortcomings of existing techniques. More specifically, our proposed approach focuses on an EfficientNetV2-B4 with extra dense layers. An end-to-end training architecture is used by the customized EfficientNetV2-B4 to calculate the deep key points and classify them in the corresponding classes. A standard dataset, specifically the Plant Village data from Kaggle, as well as real samples were taken with a drone is utilized for performance evaluation, which is complicated in the aspect of varying image samples with diverse image capturing conditions. The empirical results point out that the model outperforms several existing methods in terms of average accuracy of more than 98% in prediction. The results also demonstrate that our strategy is less time-intensive than other previous strategies while also confirming its resilience.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Detection of Pathogens in Plant Leaves Using Drone-Based Deep Learning Approach


    Weitere Titelangaben:

    Transactions on Computer Systems & Networks


    Beteiligte:
    Karthikeyan, P. (Herausgeber:in) / Kumar, Sathish (Herausgeber:in) / Anbarasu, V. (Herausgeber:in) / Remya, S. (Autor:in) / Jeyakrishnan, V. (Autor:in) / Karunakaran, V. (Autor:in)


    Erscheinungsdatum :

    27.09.2023


    Format / Umfang :

    12 pages





    Medientyp :

    Aufsatz/Kapitel (Buch)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch





    Real-Time Drone Detection Using Deep Learning

    Sricharan, K. / Venkat, M. | Springer Verlag | 2023

    Freier Zugriff



    Deep Learning-Based Autonomous Drone(s) Assistance

    Prithvi Krishna, C. / Bharatiraja, C. | Springer Verlag | 2021