An improved machine learning algorithm for predicting blast disease in paddy crop

Rice is the most important staple food crop in India. However, it is one of the crops which tends to be affected frequently by disease causing agents resulting in reduced yield. Although several challenges affect the growth of the crops, such as pest, environmental conditions and natural diseases, crop diseases is one of the major threats to food security. The diseases are caused by bacteria or fungi that can affect the crop at any stage. In the proposed system, the preprocessed images of both infected and healthy rice crops are given as input to the machine learning algorithms via., Convolutional Neural Network (CCN) for feature extraction and Support Vector Machine (SVM) for classification. The results show that CNN combined with SVM based classification method provides better accuracy than that of SVM applied alone for both infected and normal images of paddy blast disease. The proposed method would immensely benefit the farmers in terms of increased revenue and ensures food security to the nation at large. (C) 2019 Elsevier Ltd. All rights reserved.