TOMATO LEAF DISEASE DETECTION USING DEEP LEARNING

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.