An Approach to Predict Paddy Crop Disease Using Convolutional Neural Network

Abstract

In the recent past, due to excessive use of human-made wastage and pesticides, plant disease increased at a higher rate. These diseases can be dangerous in a later stage if it is not taking into account. Also, due to a lack of technical inefficiency, sometimes it becomes hard to detect these diseases. So, this paper discussed a model for detecting the disease present in rice crops. We used image processing with a Deep Learning model to specify the affected rice plant. As paddy field disease follows the same pattern, we can discriminate affected rice plant from the healthy plant. Therefore, we can detect these affected plants using deep learning methods with convolutional neural networks (ConvNet/CNN). So, we take the image of the affected plant and dynamically analyze the images of the disease. This system performs diagnose with the dataset of images using deep learning. Besides, we emphasize on the pattern created by the Bactria that reduced the learning time of the model. Thus, the system has obtained accuracy over 90% in detecting the affected corps.