Classification of Rice Leaf Disease using Convolutional Neural Network

Abstract

This research investigates the classification of rice leaf diseases using Convolutional Neural Networks (CNNs) to achieve higher accuracy than traditional transfer learning approaches. We address the challenge of accurately identifying various rice diseases from leaf images, crucial for timely intervention and yield optimization. A novel CNN architecture is designed and trained on a comprehensive dataset, demonstrating improved classification accuracy and robustness compared to pre-trained models. The results highlight the potential of tailored CNNs for precise and efficient rice disease diagnosis.

Introduction

Rice is a staple food crop, and its production is significantly impacted by various diseases. Early and accurate disease detection is vital for effective management and preventing substantial yield losses. While traditional methods rely on expert knowledge and are time-consuming, image-based automated diagnosis offers a promising alternative. However, the complexity of visual symptoms and the need for high accuracy present challenges. Existing transfer learning methods, while useful, often lack the fine-grained feature extraction necessary for differentiating subtle disease variations. This project aims to overcome these limitations by developing a specialized CNN architecture optimized for rice leaf disease classification.

Objectives

• To design and develop a high-accuracy CNN model for classifying rice leaf diseases.
• To achieve superior classification accuracy compared to existing transfer learning methods.
• To evaluate the robustness and generalization capability of the proposed model.