Image recognition of four rice leaf diseases based on deep learning and support vector machine

Abstract

In the field of agricultural information, identification and prediction of rice leaf diseases has always been a research focus. Deep learning and support vector machine (SVM) technology are hot research topics in the field of pattern recognition at present. Their combination can not only solve the problem effectively, but also improve the recognition accuracy. In this study, firstly, we use convolution neural networks (CNNs) to extract the rice leaf disease images features. Then the SVM method is applied to classify and predict the specific disease. The optimal parameters of SVM model are obtained through the 10-fold cross validation method. The experimental results show that when the penalty parameter C=1 and the kernel parameter g = 50, the average correct recognition rate of the rice disease recognition model based on deep learning and SVM is 96.8%. This accuracy is higher than that of the traditional back propagation neural networks models. This study provides a new method for the further research of crop diseases diagnosis by using deep learning.