Advanced Rice Leaf Disease Detection Using AgriRover: Enhancing Agricultural Monitoring and Crop Health Management

Agriculture is pivotal for the economy of Bangladesh, providing livelihoods for the majority of its population and contributing significantly to its GDP. Rice stands out as the most significant crop among various agricultural products, yielding the highest volume of production in this country. A variety of diseases affecting rice crops can lead to reduced production, thereby impacting both the economy and demand. For this, early detection of these diseases can mitigate their detrimental effects on healthy crops and facilitate prompt pesticide application. This research involves the creation of a machine learning algorithm trained for identifying and categorizing healthy and diseased rice crops. Subsequently, the system is integrated into a rover that makes the system capable of traversing rugged land to assess the crops. A convolutional neural network known as DenseNet, comprising 201 layers, is introduced and leveraged as a pre-trained model for the classification of six categories of rice diseases. To ensure practicality and dependability in real-world scenarios, a rover is specifically engineered, enabling the deployment of the model. The rover is equipped to capture images of rice leaves and autonomously determine whether they are diseased or healthy, hence improving the overall detection and classification system of crops. The suggested approach represents a significant advancement in the holistic management of agricultural monitoring and crop health.