Sugarcane Leaf Disease Classification using Transfer Learning

Abstract

Agriculture is crucial to the Indian economy as it provides employment to roughly half of India's population and contributes to 17% of India's GDP. Since 1947, India has seen an enormous increase in the yield and produce of crops. Yet around 50,000 crore worth of crops are lost to pest and disease attacks every year. According to the United Nations Food and Agriculture organization, there is an approximate loss of 40% in production of crops globally due to pests and diseases. This costs the global economy more than $220 billion annually. One of the most significant cash crops grown by farmers in India is Sugarcane. Red rot and Red rust epidemics have been common for sugarcane cultivators in India. With the rise in technology and artificial intelligence, there are various methods that can provide a solution to this issue. Our paper discusses in detail about using DenseNet201, a transfer learning model, along with Support Vector Machine in the output layer to detect Red Rot and Red Rust diseases in sugarcane leaves.