Empowering Rice Farmers: Automated Multi-Classification of Rice Diseases using YOLOv5

Rice is a staple grain for a large percentage of the world's population, and agriculture plays a critical role in maintaining global food production. However, rice infections are a danger to harvest success and food safety. The proposed study uses you only look once version 5 (YOLOv5) deep learning (DL) model to create a dependable and precise method for multi-classifying rice diseases. The mission is to provide rice farmers and agricultural specialists with a reliable method of disease diagnosis and classification that will allow for prompt responses and efficient disease management. The approach utilizes a dataset of 20,000 photos of rice plants that the user collects to train the YOLOv5 model. Besides “healthy,” “bacterial blight,” “brown spot,” “eyespot,” “false smut,” and “sheath rot” are also included in the dataset's six categories. The suggested study evaluated the YOLOv5 model's efficacy using measures derived through meticulous data preparation, model training, and evaluation. The outcomes show that the YOLOv5 model has a high F1 score, precision, and recall, and an accuracy of 94.03%. The suggested method for classifying rice diseases is superior to a state-of-the-art model, as shown in the comparison results. The solution provides farmers with actionable data for prompt disease management, decreasing crop losses, and increasing agricultural output by precisely identifying and categorizing rice illnesses. This study highlights the potential of DL techniques, namely the YOLOv5 model, to completely transform the agricultural industry. The established method aids farmers in making educated decisions to safeguard crops and increase food security, providing a realistic answer to the urgent problem of rice illnesses.