Artificial Intelligence based Classification of Diseases for Rice Leaf Using CNN model

Rice cultivation is a crucial industry in India, but it is plagued by various diseases that can damage crops at different stages. These diseases are challenging for farmers to identify accurately due to their limited knowledge and expertise. As a result, the farmers often struggle to take appropriate measures to prevent or manage these diseases, which can result in significant losses in crop yield and quality. Therefore, there is a need for advanced technologies and tools to help farmers accurately identify and manage these diseases, ensuring a sustainable and profitable rice cultivation industry in India. Recent advances in Deep Learning have demonstrated that Convolutional Neural Network models can be highly effective in automatic image recognition tasks. These models have shown great potential in addressing the challenges faced by farmers in identifying diseases in crops such as rice. However, in order to train such models, a large and diverse dataset is required, which may not always be readily available. To address this issue, researchers have created their own dataset of rice leaf disease images, which may be smaller in size but sufficient for the task at hand. To develop their CNN model, they have used a technique called Transfer Learning, which used as a starting point to fine-tune already trained models for a new task. The proposed CNN architecture is based on the VGG-16, a widely used pre-trained model used in computer vision tasks. The researchers trained and tested their model with a dataset obtained from rice fields and the Internet. The results show that the proposed model achieves 92.46% accuracy, demonstrating its potential in accurately detecting rice leaf diseases.