Rice leaf disease detection using convolutional neural networks

Abstract

Rice is one of the major developed crops in Bangladesh which is influenced by different infections at different stages of its cultivation. It is exceptionally troublesome for the farmers to manually identify these infections precisely with their constrained knowledge. Recent improvements in Profound Learning appear that Automatic Image Acknowledgment frameworks utilizing Convolutional Neural Network (CNN) models can be exceptionally advantageous in such issues. Since rice leaf malady picture dataset is not effortlessly accessible, we have created our possess dataset which is little in measure subsequently we have used Transfer Learning to create our profound learning show. I use primary data which are collected from different cultivation field in Tangail. The proposed CNN engineering is based on InceptionResnetV2 and is trained and tried on the dataset collected from rice areas. The exactness of the proposed demonstrate is 93.12%.