Articles producció científica> Enginyeria Informàtica i Matemàtiques

Color-Aware Two-Branch DCNN for Efficient Plant Disease Classificat

  • Identification data

    Identifier: imarina:9280248
    Authors:
    Schuler JPSRomani SAbdel-Nasser MRashwan HPuig D
    Abstract:
    Deep convolutional neural networks (DCNNs) have been successfully applied to plant disease detection. Unlike most existing studies, we propose feeding a DCNN CIE Lab instead of RGB color coordinates. We modified an Inception V3 architecture to include one branch specific for achromatic data (L channel) and another branch specific for chromatic data (AB channels). This modification takes advantage of the decoupling of chromatic and achromatic information. Besides, splitting branches reduces the number of trainable parameters and computation load by up to 50% of the original figures using modified layers. We achieved a state-of-the-art classification accuracy of 99.48% on the Plant Village dataset and 76.91% on the Cropped-PlantDoc dataset.
  • Others:

    Author, as appears in the article.: Schuler JPS; Romani S; Abdel-Nasser M; Rashwan H; Puig D
    Department: Enginyeria Informàtica i Matemàtiques
    URV's Author/s: Abdellatif Fatahallah Ibrahim Mahmoud, Hatem / Abdelnasser Mohamed Mahmoud, Mohamed / Puig Valls, Domènec Savi / Romaní Also, Santiago / Schwarz Schuler, Joao Paulo
    Keywords: Plant disease Neural networks Multipath Deep learning Dcnn Cnn Cie lab Artificial intelligence
    Abstract: Deep convolutional neural networks (DCNNs) have been successfully applied to plant disease detection. Unlike most existing studies, we propose feeding a DCNN CIE Lab instead of RGB color coordinates. We modified an Inception V3 architecture to include one branch specific for achromatic data (L channel) and another branch specific for chromatic data (AB channels). This modification takes advantage of the decoupling of chromatic and achromatic information. Besides, splitting branches reduces the number of trainable parameters and computation load by up to 50% of the original figures using modified layers. We achieved a state-of-the-art classification accuracy of 99.48% on the Plant Village dataset and 76.91% on the Cropped-PlantDoc dataset.
    Thematic Areas: Theoretical computer science General computer science Decision sciences (miscellaneous) Control and systems engineering Computer science (miscellaneous) Computer science (all) Computational mathematics Artificial intelligence
    licence for use: https://creativecommons.org/licenses/by/3.0/es/
    Author's mail: mohamed.abdelnasser@urv.cat hatem.abdellatif@urv.cat joaopaulo.schwarz@estudiants.urv.cat santiago.romani@urv.cat domenec.puig@urv.cat
    Author identifier: 0000-0002-1074-2441 0000-0001-5421-1637 0000-0002-7582-0711 0000-0001-6673-9615 0000-0002-0562-4205
    Record's date: 2024-09-21
    Papper version: info:eu-repo/semantics/publishedVersion
    Link to the original source: https://mendel-journal.org/index.php/mendel/article/view/176
    Licence document URL: https://repositori.urv.cat/ca/proteccio-de-dades/
    Papper original source: Mendel. 28 (1): 55-62
    APA: Schuler JPS; Romani S; Abdel-Nasser M; Rashwan H; Puig D (2022). Color-Aware Two-Branch DCNN for Efficient Plant Disease Classificat. Mendel, 28(1), 55-62. DOI: 10.13164/mendel.2022.1.055
    Article's DOI: 10.13164/mendel.2022.1.055
    Entity: Universitat Rovira i Virgili
    Journal publication year: 2022
    Publication Type: Journal Publications
  • Keywords:

    Artificial Intelligence,Computational Mathematics,Computer Science (Miscellaneous),Control and Systems Engineering,Decision Sciences (Miscellaneous),Theoretical Computer Science
    Plant disease
    Neural networks
    Multipath
    Deep learning
    Dcnn
    Cnn
    Cie lab
    Artificial intelligence
    Theoretical computer science
    General computer science
    Decision sciences (miscellaneous)
    Control and systems engineering
    Computer science (miscellaneous)
    Computer science (all)
    Computational mathematics
    Artificial intelligence
  • Documents:

  • Cerca a google

    Search to google scholar