Machine learning approaches towards sugarcane crop disease identification

Abstract

Sugarcane is an important crop which plays a vital role in overall revenue in agriculture. Almost many of countries in world are involved in sugarcane plantation and pests affect sugarcane yield or production. The use of decision support system in agriculture for crop disease identification is made possible using machine learning algorithms. We evaluate the performance of machine learning techniques for the sugarcane crop in this research. We focus on mainly four diseases that affect of yield of sugarcane crop and we train machine algorithms using the images of these crop diseases using CNN, CNN and SVM, XGBoost algorithm. The performance of these algorithms for the classification is evaluated as to how it synthesizes the features and correctly classifies the disease.