Rice Plant Disease Detection using Convolutional Neural Networks

Accurately identifying paddy diseases is essential for achieving optimal and quality yield. However, the traditional method of identifying diseases depends on human expertise, and it is prone to errors. In this paper, we use Convolutional Neural Networks (CNNs) and deep learning approaches to identify various rice plant diseases like blast, brown spot and bacterial blight. The CNN model is trained on images of different plant diseases, and various models are evaluated to determine the most effective one for disease identification. The findings of this research will contribute to automated paddy disease diagnosis, aiding farmers in timely and effective disease management. The various models attained different accuracies in paddy disease classification: DenseNet121 achieved an accuracy of 97.50%, the Xception algorithm achieved 96.32%, EfficientNet B4 achieved 96.25%, and MobileNet V3 Large also achieved 96.25%.