Real-Time Chili Harvest Estimation with Object Detection Technology in Enhancing Agricultural Efficiency

Abstract

Agriculture plays a critical role in ensuring food security, yet technological advancements in this sector remain limited compared to its potential. Modern innovations, such as Object Detection, offer promising solutions to enhance agricultural efficiency and productivity. This study explores the application of the YOLOv8 algorithm, the latest evolution of the YOLO object detection framework, to detect chili fruits rapidly and accurately. By employing this method, farmers can estimate potential chili harvests, streamlining yield prediction and improving decision-making processes in real time. Experimental results demonstrate that the model achieved a mean Average Precision (mAP) of 75.4%, an F1 score of 77.21%, a Precision of 74.7%, a Recall of 79.9%, and a processing speed of 6.9 
milliseconds per image. These results highlight the model's effectiveness in practical applications but also indicate room for improvement, as performance is influenced by the limited number of training iterations. Future work could focus on increasing training iterations and expanding the dataset to enhance detection accuracy and robustness, ultimately supporting precision 
agriculture advancements.