TOMATO LEAF DISEASE CLASSIFICATION USING TRANSFER LEARNING

Abstract

— Crop diseases are a vital threat to food security, but their quick identification remains arduous in
various parts of the world due to the lack of necessary infrastructure. In the literature, different methods of
plant leaf disease detection have been used. These methods were time-consuming and could not cover large
areas for the detection of leaf diseases. As deep convolution neural networks (DCNN) and transfer learning
has been favorably implemented in many fields, it has freshly moved in the domain of just-in-time crop
disease detection. In this project, we are using transfer learning to design a model that can achieve maximum
accuracy. An innovative way of training and methodology was used for quick and easy implementation of
the system in practice. The developed model was capable of detecting different plant leaf diseases out of
healthy leaves.