Fast and Precise: YOLO-based Wheat Spot Blotch Recognition

Wheat spot blotch disease is a major threat to global wheat production and causing large grain yield losses Thus, early detection and precise assessment of illness signs are critical for effective therapy and prevention methods. In this article, we present a novel approach for wheat spot blotch disease recognition based on the You Only Look Once (YOLO) object detection model and compare its performance to two commonly used object detection models: Single Shot Detector (SSD) and Faster R-CNN. The YOLO model’s single-stage architecture provides advantages in terms of real-time processing and efficiency. We trained and tested these models’ using images of wheat leaves affected with spot blotch disease. Performance metrics such as average precision, recall and intersection over union (IoU) were used to evaluate the models' ability to detect and localize disease signs. Our findings show that the YOLO model beats both SSD and Faster R-CNN in terms of detection accuracy (98.6%) and computational efficiency, making it a promising method for automatic wheat spot blotch disease identification. This research helps to develop robust and scalable technologies for early disease detection in wheat crops, allowing for timely intervention and, eventually, ensuring global food security.