A Comprehensive Review of Deep Learning-Based Approaches for Rice and Wheat Leaf Disease Detection

Abstract

Agricultural productivity is greatly influenced by various plant diseases, especially those affecting essential crops like rice and wheat. Timely and precise identification of these diseases plays a critical role in maintaining food security and enhancing crop yields. Traditional methods of detecting diseases often involve manual inspection, which is time-consuming, labor-intensive, and prone to errors. To address these challenges, there has been a growing interest in automated disease detection systems that leverage deep learning and computer vision techniques. This study focuses on utilizing image processing and machine learning to identify and classify diseases on the leaves of rice and wheat plants. The proposed system relies on a collection of images representing both healthy and diseased leaves, which are pre-processed by removing noise, enhancing contrast, and extracting relevant features. Various deep learning models, including Convolutional Neural Networks (CNNs), are employed to accurately detect and categorize the diseases. For rice, the most prevalent diseases include bacterial blight, brown spot, and blast, while wheat faces threats from diseases such as rust, powdery mildew, and leaf blight. The performance of the model is evaluated using metrics such as accuracy, precision, recall, and F1-score. Additionally, the model’s robustness is assessed under varying lighting conditions, leaf orientations, and background noise. Challenges such as small datasets, diseases with similar symptoms, and the application of the same model in different climates are discussed. The findings suggest that deep learning models can significantly improve disease detection accuracy compared to traditional methods, providing farmers and agricultural experts with an efficient, cost-effective, and scalable solution.