A Machine Intelligent Framework for Detection of Rice Leaf Diseases in Field Using IoT Based Unmanned Aerial Vehicle System

Abstract

Rice is an important food in our day-to-day life. It has rich sources of carbohydrates that are highly essential for body growth and development. Rice is an important crop in agriculture, where it enhances a country’s economy. However, if rice plants arediseased and not monitored regularly then the crop in the field is wasted and it reduces the proper production rate. Therefore, there should be a mechanism which regularly monitors the crop in a field to detect any disease to rice plant. In this paper, a framework is proposed for identification of rice leaf disease using IoT based Unmanned Aerial Vehicle (UAV) system. Here, the UAV monitors an entire field, capture the images and sends the images to the machine intelligent cloud for detection of rice leaf diseases. The cloud is installed with a proposed stacking classifier that classify the diseased rice plant images received from UAV into different categories. The dataset of these rice leaf diseases is collected from Kaggle source. The performance of the stacking classifier installed at the cloud is evaluated using Python based Orange 3.26 tool. It is observed form the results that stacking classifier outperforms the conventional machine learning models in detecting the actual disease with a classification accuracy (CA) of 86.7%.

Efficient Damage Assessment of Rice Bacterial Leaf Blight Disease in Agricultural Insurance Using UAV Data

“…In Indonesia, an objective evaluation method acceptable to insurance subscribers is required because the results of the visual damage evaluation method may differ between assessors [10,11]. Recently, there has been a growing expectation to build more robust and efficient methods for detecting paddy rice diseases using machine and deep learning, as automated approaches to detect leaf diseases can help farmers detect diseases with or without human intervention [12][13][14][15].…”