Published Paper Details:

ABSTRACT Rice is one of the most important crops in the world, feeding more than half of the global population. However, rice plants are highly vulnerable to diseases, and if these are not detected early, they can lead to serious crop losses. Traditionally, farmers identify diseases by visually inspecting leaves, but this process can be slow, inconsistent, and error-prone--especially in the early stages. To solve this problem, this work presents an automated and efficient system for detecting rice leaf diseases using deep learning. The model is built using a lightweight Convolutional Neural Network (CNN), specifically MobileNetV2, which is known for its speed and low computational requirements. The system is trained on a dataset containing three common rice diseases: Narrow Brown Spot, False Smut, and Leaf Smut. To improve performance, image preprocessing and data augmentation techniques are applied, allowing the model to handle variations in lighting, orientation, and background. In addition to predicting the disease, the system also provides treatment suggestions, making it more useful as a practical tool for farmers. The model achieves an accuracy of around 75-80%, with good precision, recall, and F1-score. Due to its lightweight design, it can run efficiently on devices with limited resources, such as smartphones. A simple web interface is also developed, enabling users to upload images and receive instant results. Overall, this system provides a fast, scalable, and practical solution for crop disease detection and has the potential to be extended to more diseases and real-time agricultural applications..