LEAF DISEASE PREDICTION USING CONVOLUTIONAL NEURAL NETWORK
Abstract
Agriculture suffers greatly from leaf diseases, which result in significant yield losses. In order to effectively manage and prevent many disorders, early and correct identification is essential. Deep learning methods, such Convolutional Neural Networks (CNN), have demonstrated encouraging performance in picture classification tasks in recent years. A CNN-based methodology for predicting leaf disease is proposed in this paper. The framework is composed of three basic steps: disease categorization, feature extraction using CNN models that have already been trained, and picture preprocessing. The suggested method makes use of CNN models' capacity to automatically extract discriminative characteristics from input photos. A collection of leaf images with various disease classes is used to train and assess the model's performance. The tests show that the CNN-based framework outperforms conventional machine learning techniques in disease categorization, achieving high accuracy. With the use of this method, leaf illnesses can be predicted quickly and accurately, giving farmers the opportunity to take prompt action to manage the disease. Because CNN models are used, the system may be scaled and adjusted to fit different crop kinds and disease groups. The keywords for this abstract are: leaf diseases, Convolutional Neural Networks, image classification, deep learning, feature extraction, disease prediction, agriculture, and disease management.