Diagnosis of Tomato Plant Diseases Using Pre-Trained Architectures and A Proposed Convolutional Neural Network Model

Tomato is one of the most important vegetables in world. Presence diseases and pests growing area significantly affect choice variety tomato. Early-stage diagnosis plays an role determining whether tomato subject to effective economical chemical, mechanical biological controls, internal external quarantine. In this study, deep learning was used diagnose some tomatoes. For purpose, a novel CNN-based approach Keras models including DenseNet201, InceptionResNetV2, MobileNet, VGG16 architectures were used. Early, middle, late stages 18.456 images Bacterial Spot, Early Blight, Leaf Mold, Septoria Target Mosaic Virus, Yellow Curl Virus healthy leaves examined. The experimental results showed that custom CNN model produced 99.82% training accuracy. We recommend terms monitoring diagnosing diseases. obtained with study can be robotic spraying harvesting operations.