Sugarcane Plant Disease Detection Using Deep Learning

Abstract

Bangladesh is an agricultural country and the economy of Bangladesh is dependent on agriculture. Plant diseases are responsible for major economic losses in the agricultural industry worldwide. Sugarcane plant disease is one of them. Sugarcane is the main source of sugar and ethanol. Sugarcane plant disease is mostly known problems in our farmers. In this research, we work on various diseases of this plant to find out the solution and to have some better ways to reduce it. This research helps the farmers to know the diseases of sugarcane. In the first part of the research we collect some related papers, articles and we study about those. Then we collect images of sugarcane for datasets. Collecting images are both disease and healthy plants of sugarcane. We collected almost 2212 sugarcane images. Then we applied Convolutional Neural Network (CNN) as the basic deep learning method in our processed dataset. This helps to classify and detect the disease images. In this work Convolutional Neural Networks (CNN) achieve a best accuracy of 99.46 %.