Potato Diseases Detection Using Deep Learning Techniques

Abstract

Although potatoes aren't the most important crop in the world, they're nonetheless essential to the diets of many people. However, crop quality and production are constantly threatened by leaf diseases, which result in substantial financial losses. The wide variety of viruses that can infect potatoes makes it challenging to diagnose infections using conventional methods. In this study, we provide PLDIDL, a novel deep learning-based approach for disease diagnostics in potato leaves. Our technology offers researchers and farmers the most efficient and precise automated method for detecting potato diseases by applying state-of-the-art deep learning algorithms. Other diseases that can affect potato leaves are also included in this section. These include early blight, late blight, blackleg, and others. They are polar opposites in every way, including look and trajectory. Intricate patterns and characteristics can be extracted from leaf pictures using convolutional neural networks (CNNs). The algorithm's accuracy percentage of 98.44% when comparing healthy and diseased leaves is quite remarkable. The agricultural community has a lot of space for the proposed paradigm to flourish; it will allow for sustainable potato cultivation that can resist emerging plant diseases.