Comparative Analysis of Deep Learning Techniques for Rice Leaf Disease Detection: Object Detection and YOLO Model Evaluation

Disease categorization is important because it helps farmers apply the correct insecticide to the affected section of the plant. Studies show that rice farmers face many challenges when trying to classify diseases in their rice leaves. Incorrect categorization can lead to the use of incorrect insecticides, which worsens the situation. Taking into account the above analysis, rice farmers clarify the effects or symptoms of the disease through identification with the naked eye, as this requires time and may involve specialists in the field when distinguishing the classes. Occasionally samples need to be taken to a laboratory. This process is quite expensive and even more time consuming. For these diseases, accurate and reliable detection techniques are crucial. The study aims to improve the identification of rice leaf disease identification through the use of YOLOv9, an object recognition model. For object identification and image localization in real-time, the YOLOv9 model is employed.The study employed a merged dataset that featured photos of rice leaves afflicted with various diseases, including bacterial leaf blight, healthy, leaf scaled, leaf smut, and tungro. This study proposes using the YOLOv9 object recognition model to automate and enhance the identification of rice leaf diseases with accuracy rate of 99.5%, aiming for faster, more accurate, and cost-effective detection in real-time.