Early and Late Blight Disease Detection in Potato

Abstract

Around 125 countries cultivate the potato, which is the most significant tuber crop in the world. After rice and wheat, potato is primarily eaten by billions of individuals worldwide practically every single day. However, several diseases caused by fungi and bacteria are affecting potato crop yield and declining its quality. Because of differences in environmental conditions, detection of diseases in plants at the early stages is a difficult undertaking. Farmers now face a serious threat from two of the major prevalent potato diseases, Early blight (produced by the fungus Alternia solani) and Late blight (produced by the bacterium Phytophthora infestans). Due to these diseases the farmers are suffering from major economic losses incurred from significant crop wastage. The majority of conventional methods for diagnosing diseases are pricy, time-consuming, sample-intensive, and require skilled workers. This model's main goal is the quick and simple diagnosis of early and late blight infections in potatoes. Convolutional Neural Network (CNN) is the proposed deep learning technique in this paper. Here, a classification model based on the images is developed that can classify photos of potato leaves in their early, late, or healthy stages.