Comparison of the Use of YOLOv11 Variations in the Empty Parking Spaces Detection System

The development of a smart parking system using the You Only Look Once (YOLO) model has improved the efficiency of parking management by providing real-time vehicle detection and availability of parking spaces. This study compared three variations of YOLOv11-Nano (YOLOv11n), YOLOv11-Small (YOLOv11s), and YOLOv11-Medium (YOLOv11m) to determine the most effective model in detecting empty parking spaces. The experiment was carried out using a dataset consisting of 5725 images of parking areas with various conditions such as angles, lighting, and distance. In addition, the researcher also used a 6-second parking lot timelapse video for the test material of the model that had been trained. The results show that each variation of YOLOv11 has its own advantages in terms of accuracy, speed, and computing efficiency. YOLOv11n offers faster detection with lower resource consumption, while YOLOv11m provides higher accuracy with longer processing times. The findings of this study aim to help select the optimal YOLOv11 variant for smart parking implementation, thereby improving efficiency and accuracy in real-world applications.