Approximately half of the population of the world consumes rice every day as their daily diet. In Asia, rice is the main ingredient of the meal. However, various diseases influence the production of rice being harmed. Many factors are responsible for the issue such as climate change, pathogens and excessive usage of pesticides. Varieties of leaf diseases are there that may harm quality of rice production. To protect the crop from diseases, farmers are relying on their eyes to identify diseases. Most often human eye can possibly make mistake identifying the disease. In consequence, harvested rice gets harmed both in quality and in quantity as well. To prevent those diseases, researchers have found cutting-edge techniques such as deep neural network with image processing based frameworks to detect and classify rice plant diseases to increase crops production and to reduce losses. Recent works present all the possible techniques related to rice plant disease classification and detection. This work represents a state-of-art in disease identification approach for rice leaf that can efficiently classify several types of diseases analyzing the leaf images. By adopting the efficiency of deep learning models with hyperparameter tuning, the methodology achieved accuracy of 97.10% with ResNet50. Importantly, the experimental outcomes of the proposed framework highlight the remarkable performance comparing the results with existing state-of-the-art technique. The aim of this research is to provide a more promising future for crop management through increased productivity while addressing the urgent problem of rice plant disease detection.
Deep Learning-based Image Classification Framework for Rice Leaf Disease Detection: A Systematic Implementation Approach with Comparative Analysis
06.11.2024
623557 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch