Paddy Plant Disease Classification and Prediction Using Convolutional Neural Network

India and other Asian countries have been facing challenges over couple of decades which affect the yield in paddy plants. Challenges listed are dropping water resources/rainfall, lack of innovative scientific methods and diseases affecting the plants. These challenges have direct impact on yield of the crop eventually affect country's economy. Paddy leaf is the main investigative part of this plant providing information regarding the health status of the plant, which affect quality and quantity of the paddy crop yield. This work proposes an innovative solution with image processing and convolution neural network (CNN) to classify paddy plants as belonging to type of disease classes based on data obtained from agricultural image database repositories. CNN is natural choice as it is deep learning algorithm associated with faster convergence, accuracy in classifications with minimal training sets. This work proposes 10 layers of CNN for classification with maximum epochs of 20 with learning rate of 0.0001. Performance parameters obtained are accuracy as 94.12% and confusion matrix. Results of classification are migrated to IOT makers website.Results can provide fast and accurate safety measures such as recommendations of pesticides to arrest growth of disease. Thus system results will enhance yield of the crop.