Automated Areca Nut Disease Diagnosis: A CNN and SVM Based Model with a Bilingual Farmer Support System

Many Karnataka farmers rely heavily on areca nut farming for their livelihoods, but leaf diseases still result in significant losses in crop yield and revenue. The bilingual (Kannada/English) web application is created to assist farmers in identifying diseases early on by using images of diseased leaves. Convolutional Neural Networks (CNN) and Support Vector Machines (SVM), two potent machine learning techniques, are combined in the system. Using a locally gathered dataset of photos of areca nut leaves. The CNN was trained for 50 epochs, obtaining 90% validation accuracy and 99% training accuracy with little overfitting. Although there was some confusion with similar diseases, the SVM further increased classification accuracy. As it supports real-time implementation on edge, this system facilitates the development of a sustainable agricultural ecosystem by ensuring that local farmers can treat plants with pinpoint accuracy and precision, as well as crop-specific diseases in rural Karnataka.