DETECT RICE LEAF DISEASE USING DEEP LEARNING

Abstract

Our society is getting more and more technology-dependent day by day. Nevertheless, agriculture is imperative for our survival. Rice is one of the primary food grains. It provides sustenance to almost fifty percent of the world population and promotes a huge amount of employment [1]. Hence, discovery and prevent disease on rice are huge of importance. Which helps enhance rice production. In this thesis, we will be suggested a resolution by the mobile application to uncover the disease using deep learning. Specifically, it used EfficientNet that is a deep learning network for classification. In addition, it has utilized the pre-train model on imageNet to transfer learning on the dataset. The dataset has been collected on the Internet and amount the dataset has been collected at Rice Disease Image Dataset – Kaggle [2]. The dataset consists of five types such as brown spot, hispa, leaf blast, healthy, and others. When training is finished, it will be converted to flite format that can use on mobile devices. The model accuracy is 95% on the tsetse.