Abstract:
Automatic classification process in images has been used in many fields, especially agriculture
and medical fields in recent years. Especially in our country, image processing studies are needed
to improve agriculture and increase productivity in agriculture. Diseases in plants are a problem
that often occurs in agriculture. The disease can be caused by pests or maintenance errors. This
resulted in a decrease in agricultural production. The decline in production has resulted in a decline
in economic yields produced by farmers. Diseases of tomato plants often appear on the leaves.
In this study 10 types of leaf disease were used. In this study, the major tomato leaf diseases that
significantly affect tomato efficiency were examined and the convolutional neural networks deep
learning methods was applied for the automatic classification of these diseases. It is thought that the
model applied in this study can also be applied on other agricultural crops, so that the contribution
of image processing to agriculture will increase gradually. The model developed in this work uses
deep learning techniques: ResNet152V2 and MobileNetV2. ResNet152V2 achieved an accuracy
of 98.21% for ten class classification using images and MobileNetV2 achieved an accuracy of
91.69%. However, the best results were obtained by applying ResNet152V2 method. It can be
concluded that all the architectures performed better in classifying the diseases when trained with
deeper networks on images. The performance of each of the experimental studies reported in this
work outperforms the existing literature.
