Hybrid deep learning technique for Identifying Diseases in Sugar-cane Crops

Abstract

Sugarcane, the primary source of sugar and ethanol, is a vital crop globally. An ongoing challenge in the sugar industry is the presence of sugarcane diseases, leading to the eradication of infected crops. With-out early treatment and diagnosis of these diseases, small-scale farmers face significant financial losses. This study aimed to address the in-creasing prevalence of diseases and farmers' limited knowledge of dis-ease diagnosis and recognition. The use of deep learning techniques such as computer vision and machine learning proved to be promising. By utilizing a dataset of 13,842 sugarcane images featuring both dis-ease-infected and healthy leaves, a deep-learning model was trained and tested, achieving an accuracy rate. The trained model successfully met its objectives, and the research was finally submitted to Conven-tional neural network (CNN), Recurrent neural network (RNN).and other related models for further consideration.