A novel hybrid cnn-rnn model for sugarcane disease identification in agricultural fields

Abstract

The world's most important crop is sugarcane, which is the main source of both sugar and ethanol. The existence of sugarcane diseases, which result in the removal of afflicted crops, is a persistent problem in the sugar business. Small-scale farmers risk suffering large financial losses if these diseases are not identified and treated early. The growing incidence of illnesses and farmers' inadequate understanding of disease diagnosis and identification were the focus of this investigation. The application of deep learning methods, including machine learning and computer vision, showed promise. A deep-learning model was trained and evaluated using a dataset of 13,842 photos of sugarcane that included both diseased and healthy leaves, and it achieved an accuracy rate. The research was ultimately submitted to recurrent neural networks (RNN), conventional neural networks (CNN), and other similar models for additional evaluation after the trained model effectively achieved its goals.