AI-Driven Potato Disease Classification: ResNet50 for Detecting Late Blight in Crop Leaves

A terrible plant disease, potato late blight seriously lowers potato crop output, therefore endangering world food security. Using the ResNet50 deep learning architecture, we aim in this work to classify potato leaves into two categories: healthy and diseased with late blight. Early on identification and classification of potato leaf health is essential for both sustainable agriculture and prompt disease control. Following a 70:30 split ratio, 430 high-resolution photos make up the dataset used in this work; 302 images are used for training and 128 images for validation. JPG format makes all images consistent, thereby guaranteeing consistent preprocessing. Using transfer learning to capitalize on pre-trained ImageNet weights, the ResNet50 model was optimized for this classification challenge. With an accuracy of 98% on the validation set, our method showed amazing performance, therefore demonstrating the model's ability to differentiate between healthy and diseased leaves. These findings highlight the possibilities for machine learning models such as ResNet50 to be included into automated crop monitoring systems, so encouraging responsible consumption and production methods and helping early identification of potato late blight, so contributing to zero hunger and improved agricultural output.