Sugar Learning: Deep Learning for Rapid Detection and Classification of Sugarcane Diseases
Simrat Kaur Brar, Rishabh Sharma, Satvik Vats, Vinay Kukreja
Abstract
Sugarcane is a widely cultivated crop due to its high demand and supply in various industries. However, the increase in production levels has led to an increase in the number of diseases that affect the crop. One of the most devastating diseases is sugarcane red rot (SRR) disease. A multi-layer perceptron (MP) based deep learning (DL) model was built for the identification and classification of SRR illness using a dataset of 20,000 photos of sugarcane leaves in order to address this problem. This model was trained on 20,000 photographs of sugarcane leaves. The dataset was classified into 5 different disease severity levels. The proposed model achieved an accuracy rate of 97.97% for binary classification and an accuracy rate of 98.03% for overall multi-classification. Furthermore, a comparison of the different SRR illness stages was carried out, and it was demonstrated that the proposed model is an effective tool for accurately categorizing sugarcane images based on the severity of SRR disease. This study contributes to the development of an efficient and accurate model for the early detection and diagnosis of SRR disease in sugarcane crops, which is essential for improving crop yield and preventing economic losses.