Deep Learning-Based Classification of Tomato Leaf Diseases: A Comparative Review of Approaches

Abstract

Plant diseases cause to decreases in product in agriculture. The majority of farmers find it challenging to identify and control plant diseases. It is necessary to spot diseases early in order to minimize future losses. In this paper, a deep learning method utilizing convolutional neural networks (CNNs) to identify tomato leaf diseases is described. For example, to classify tomato leaf images into one of ten categories—healthy, yellow leaf curl virus (YLCV), bacterial spot (BS), early blight (EB), leaf mold (LM), spectoria leaf spot (SLS), target spot (TS), two spotted spider mite spot (TSSMS), mosaic virus (MV), and late blight (LB)—the proposed method initially prepares the images of the leaves. A dataset which include 16021 images of tomato leaves was used to train the model. After 10 epochs, 20 epochs, and 50 epochs of training, the accuracy was 64%, 94%, and 97%, in that order. The results provided show that the proposed approach has been effective for determining tomato leaf diseases, and that its performance gets improve with time. The early detection and prevention of tomato diseases by the use of an automated method has the capacity to boost tomato crop quality and output. The In this review paper we are going to show the different approaches that can be used to deal with the classification of tomato leaf diseases.