POTATO DISEASE CLASSIFICATION USING DEEP LEARNING

Abstract

A commonly cultivated crop is the potato. Potatoes are an essential food rich in carbohydrate serving as a staple food, thus being highly prioritized to be grown for creating a robust food security system to support successful potato cultivation. Nonetheless, potato is affected by number of diseases that hinder their development and interrupts our entire agricultural support system. Thus, early disease diagnosis can offer a better means of ensuring successful crop cultivation. Our goal is centered on a deep learning system based on K-means clustering segmentation that identifies and categorizes illnesses of the potato leaf. The network models that we have chosen are VGG16, 2-D CNN and ResNet50. Out of these chosen models, we have obtained high accuracy using VGG16, provided optimum result among three networks. The model's performance is compared with the existing state-of-the-art methods for enhanced accuracy while working on large dataset created from multiple resources