Abstract:
Good yield from banana farms always depend on healthy and disease-free leaves of banana. Hence, it is very essential to detect diseases on time for proper precautions. Ma...Show MoreMetadata
Abstract:
Good yield from banana farms always depend on healthy and disease-free leaves of banana. Hence, it is very essential to detect diseases on time for proper precautions. Manual detection and classification of diseases costs large amount of time and experts' involvement. But implementation of an automated system can help this process within no time. This paper presents three models for banana leaf disease detection and classification using two machine learning approaches, KNN and SVM, and a deep learning approach Alexnet. The Leafspot and Sigatoka are the diseases detected and classified in this work. The RGB colour images are used to train the model to detect and classify the diseased and healthy leaves with and without background. The preprocessed images after data augmentation are used for training the model. The algorithms gave testing accuracies of 76.49%, 84.86% and 96.73% for KNN, SVM and Alexnet respectively.
Date of Conference: 24-26 November 2022
Date Added to IEEE Xplore: 16 February 2023
ISBN Information: