Hybrid Neural Network and Evolutionary Model for Detection of Rice Plant Disease

Rice production is still the most important thing for food needs in Indonesia. However, rice plants cannot be separated from pests and diseases, especially diseases in rice plants. Rice plant diseases can affect the amount of rice production. From these problems, we need a method that can detect rice plant diseases. In this study, we propose a hybrid model that combines Neural Network and Evolutionary Algorithm models and feature extraction of Gray Level Co-Occurrence Matrix (GLCM) to detect rice plant diseases. In this study, the evolutionary algorithm uses Genetic Algorithm (GA). The proposed method can determine the value of Neural Network parameters including epoch, learning rate, and momentum which can provide the best accuracy value. The image dataset of rice plant disease is used in this study. The experimental results show that the accuracy of the proposed method produces an accuracy of 97.5%. From the results of research on rice plant diseases, this study produces the best accuracy compared to previous studies.