Effectiveness of Deep Learning Model for Plant Disease Detection

Abstract

In recent years, agriculture has become more and more important since global warming became a serious problem human beings must face. In this case, plant health needs to be noticed. It is very important to avoid food shortages as much as possible, given the fact that food is still scarce in today’s society. If plants have any disease, the earlier people find it, the easier for the farmers to carry out the action to stop the disease and protect plants. Compared to the traditional machine learning Convolutional Neural Network (CNN) , deep learning has become very efficient dealing with the data. In this paper, two deep learning, specifically two convolutional neural network models, are compared to two machine learning models, on a plant disease diagnosis dataset. The author finds out the deep-learning-based model performs superior than traditional machine learning models. Moreover, designing domain specific techniques for the plant disease detection would help the accuracy of a model.