Deep Learning in Agriculture: Detection and Analysis of Sugar Beets with YOLOv8

In this study, the performance of the YOLOv8 model in detecting sugar beets was evaluated using images obtained from a drone over a sugar beet field. High-resolution drone images were divided into small segments, labeled, and the model was trained using data augmentation techniques. The results obtained during the training and testing phases demonstrated that the model successfully detected sugar beets with high accuracy, precision, recall, and F1 score values. The analysis of label correlograms and result graphs confirmed the model’s labeling accuracy and detection capability. These findings indicate that the YOLOv8 model can be an effective tool in agricultural production monitoring and plant health assessment applications. In the future, the model’s performance will be more comprehensively evaluated using datasets obtained from different geographical regions and various agricultural products.