A detecção de doenças em folhas do tomateiro

Abstract

Tomatoes (Solanum lycopersicum) are a rich source of essential nutrients, such as vitamin C and lycopene, playing a crucial role in the population's diet and local economy. This study arises from the need to address specific challenges in the cultivation of these fruits, focusing on the identification of foliar diseases through deep learning models. Utilizing a database with 11,000 images of tomato leaves, various classes of pathologies were explored, including Bacterial Spot, Black Spot, Late Blight, Leaf Mold, Septoria Leaf Spot, Two-Spotted Spider Mite, Target Spot, Yellow Leaf Curl, Mosaic Virus, along with healthy samples. Three models were trained: ResNet50, MobileNet V2, and Inception V3. The results reveal that Inception V3 stands out, achieving an accuracy of 98%, followed by ResNet50 with 96%, and MobileNet V2 with 95%. These findings emphasize the consistency of Inception V3 in classifying tomato leaf images. From the best model, an API was developed using the Django Rest Framework, capable of processing images and providing information about pests and control methods. Tests validated the effectiveness of the API, representing a significant contribution to the advancement in detecting diseases in tomato plantations