Tomato Leaf Disease Detection Using Deep Learning Based Model

Tomato yield has been significantly affected by four distinct types of leaf diseases: Alternaria leaf spot, Brown spot, Mosaic, and Grey spot. However, a reliable and rapid disease detector for Tomatoes is still lacking in the current body of research, which could threaten the industry's growth. This research suggests a deep learning strategy for immediate-form Tomato leaf disease detection using enhanced convolutional neural networks (CNNs). In this study, we first use data augmentation and picture annotation tools to build the Tomato leaf disease dataset, which includes both standardized laboratory photos as well as complicated images captured in the field. In light of this, we propose a new deep-CNN-based model for disease detection in Tomato leaves, using the MobileNet architecture. Finally, the proposed MobileNet model are trained for identifying these five prevalent Tomato leaf diseases using a dataset of 962 photos of diseased Tomato leaves for the hold-out testing dataset. The detection performance of the proposed framework is demonstrated experimentally to be 95.79% mAP. In this study, we show that the innovative model offers a high-performance solution for the early identification of Tomato leaf diseases, outperforming state-of-the-art methods in terms of detection accuracy and detection speed in real-time.