Paddy Crop Disease Detection using LeNet and MobileNet Models

The paper explores the transformative impact of automated disease detection systems in agriculture, focusing on the efficient detection of rice leaf diseases using deep learning techniques. This study emphasizes four key architectures: Convolutional Neural Networks (CNN), VGG-19, LeNet, and Mobile-Net, known for their pivotal role in the accurate classification of diverse rice leaf disease classes. The dataset used in this work, features nine disease categories and a healthy state, serves as the foundation for our investigation. The extensive study and research revealed each model's significant potential for improving agricultural disease control methods. With a remarkable accuracy of 96.1%, Mobile-Net stood out as the leader, showing its edge in precise diagnosis of diseases. While VGG-19 set a baseline with 85.4% accuracy, LeNet showed practical possibility with an accuracy of 90.2%, especially in environments with limited resources.