All of us are familiar with potatoes as a vegetable. In India, growing potatoes has become quite popular in recent years. However, diseases like early blight and late blight are impeding potato production and driving up production costs. To boost potato production and digitize the system, the goal is to create an automated and quick disease detection method. Our primary objective is to use leaf images to diagnose potato disease using the CNN algorithm. This paper provides a visual representation of how automated systems based on machine learning will identify and categorize potato leaf diseases. The most effective method for identifying and analyzing these disorders is image processing. In this analysis, picture division is carried out; over 2000 images of healthy and unhealthy potato leaf are taken from Kaggle, and a few pre-made models are used to identify and classify healthy and diseased leaves. With 30% test data and 70% train data, the algorithm successfully predicts one of them with an accuracy of 91.41% in testing. Our results demonstrate that CNN outperforms all currently available tasks for detecting potato disease. Soft computing technology is desired by plant pathologists for the precise and dependable diagnosis of plant diseases

Potato disease, late blight, early blight, CNN, image processing, V3, VGG16, VGG19, Feature extraction