ABSTRACT
In India rice is the primary Kharif crop which is cultivated in the autumn season and consumed more or less in every state of India however the crop may be vandalized by numerous diseases. In agriculture domain the automatic identification of rice disease is essential as it influence the economy as well as the food supply of the nation. Modern AI methods are appropriate choice for detection of rice diseases as it has the capabilities to extract the features which are important for the classification and remove the undesirable and destructive features. This paper presents an intelligent method for classification of rice disease among three classes. It is demonstrated in the paper that Deep Neural Network can predict the diseases more efficiently than other AI methods like SVM. The classification accuracy of 98.3%, with area under the curve 99.9% achieved by the model.
