Hybrid Deep Model for Automated Detection of Tomato Leaf Diseases

Tomatoes are preferred by farmers because of their high productivity. This fruit has a fibrous structure and contains plenty vitamins. Tomato diseases generally observed on stem, fruit, leaves. Early diagnosis the disease in plants is vital importance for plant. very important who expect economic gain from that Because if not treated early, these tomatoes should be destroyed. For reasons, systems to diagnose early important. In this study, tomato leaf classification model developed with deep learning methods, which one most popular artificial intelligence techniques, proposed order eliminate possibility human eye being mistaken. 6 different Convolutional Neural Network (CNN) architectures were used. first stage consists two stages, process was carried out Alexnet, Googlenet, Shufflenet, Efficientb0, Resnet50, Inceptionv3 previously trained. second stage, feature maps images dataset obtained using six pre-trained architectures. hybrid extracted best models concatenated. Then, Neighborhood Component Analysis (NCA) method applied features speed up system, unnecessary removed optimized. The optimized map classified traditional intelligent models. As result experimental studies, average accuracy rate 99.50 percent.