Tomato Plant Disease Detection using Deep Learning Based Techniques: A Comparative study
P. Deepika, B. Arthi
Abstract
Tomato is most widely used and important crop in India. It plays a significant role in agriculture. Tomato plant grows in short period of time yields more tomatoes. Tomato is good in both nutrition and income. It can be grown in any soil condition and temperature. There are various diseases that affect tomato. To increase the production of tomato early detection of disease is important. There are various diseases in tomato plant like bacterial wilt, early blight, Late blight, Leaf, septorial leaf spot, leaf mold, bacterial spot etc. Deep Learning technologies provides better result in disease identification and detection. Deep learning architectures AlexNet, ResNet50, DenseNet121, EfficientNetB5, Inception_Resnet_V2, InceptionV3, MobileNet, VGG16 and VGG19 are tested against tomato plant diseases. The implementation is done by taking the tomato leaf images from real environment. The nine architectures are compared and in which DenseNet121 and MobileNet provides better accuracy than other deep learning architectures.