Smart Agriculture: Detecting and Classifying Leaf Diseases with ML Techniques

Abstract: The concept of savvy cultivating is picking up pace in the agrarian segment these days, and it utilizes sensors and an extent of machine learning-based advances. As per later overviews, 56 percent of the agrarian segment is enduring overwhelming misfortunes due to illnesses shaping plant clear-outs. It's imperative to screen the spread of the infection and progress agrarian yields. We are required to recognize the malady to begin with on time in order to maintain a strategic distance from it spreading. Hence, we can address this issue by executing a few calculations to identify affliction and clear it out. In this paper, we have compared support vector machines (SVM), K-Nearest Neighbor (KNN), and convolutional neural networks (CNN). The three different models are presented and examined in this ponder, and they are competent at distinguishing eight different leaf illnesses. The CNN demonstration has come to a 96 percent precision when preparing the pictures of the soybean leaf illness dataset, outperforming the KNN and SVM models, which have exactness of 64 percent and 76 percent, respectively.