Enhanced Potato Leaf Disease Detection Using Deep Learning

Abstract

Potato crops are essential to global food security but are highly susceptible to leaf diseases such as early blight or late blight, which negatively impact yield and quality. Conventional detection approaches are time-consuming and error-prone. The paper proposes a deep learning-based framework that employs Convolutional Neural Networks (CNNs) along with complementary image processing methods to accurately identify and classify potato leaf diseases. with feature extraction based on color, texture, and morphological characteristics and hyperparameter optimization using a comprehensive dataset, the model yields above 99.99% classification accuracy with considerable precision and recall. This approach facilitates early disease detection, appropriate intervention, and minimizes crop losses, leading to sustainable agriculture and improved food security worldwide.