An Improved Soybean Foliar Disease Detection System using Deep Learning

Abstract

India facing the issue of high rises in the prices of cooking oil. And soybean oil is the second most commonly used oil for cooking in India. But due to environmental hazards like heavy rain, floods etc and also due to diseases the growth ratio of many crops including soybean is reduced. Soya plant leaf diseases are a big problem in front of all farmers. The main issue of automatic detection is resolve in this research. Soya plant plants were mainly affected by the illness such as brown spot syndrome; bacterial blight and frog eye are the most deadly diseases of soybean. These diseases, if detected early and with the important treatment measures, limit the significant economic losses to the farmers. In this study, proposed models will successfully classify and detect the soya plant leaf disease using the CNN. Diseases symptoms were extracted from soya plant leaves and these images were then processed with the CNN classification and provide the higher accuracy of 95.09%. This model successfully identifies infected foliar regions efficiently.