<![CDATA[The number of visitors to Museum Buleleng is still counted manually using a mechanical counter, which is less efficient, less accurate, and unable to provide real-time data, particularly when several visitors arrive simultaneously. This study aimed to design, implement, and evaluate a visitor-counting application based on You Only Look Once version 8 (YOLOv8) using a tracking method. The system used a webcam as the video source, YOLOv8 to detect person objects, and object ID tracking to prevent double counting. The application was developed using Python and OpenCV with a graphical user interface designed for ease of use. Testing conducted in the Public Relations Room of the Faculty of Engineering and Vocational Education at Universitas Pendidikan Ganesha (Undiksha) showed that the system achieved an average detection accuracy of 94.7% with a processing speed of 31 FPS, while the implementation of the tracking method was able to reduce counting duplication to 2.1%. During testing, the system recorded 50 incoming visitors, consisting of 22 females and 28 males, and 42 outgoing visitors, consisting of 18 females and 24 males, indicating that eight visitors remained in the room at the end of the observation. These results show that the application is capable of automatically, accurately, and in real time counting and monitoring the number of visitors. This study provides a practical contribution to the application of computer vision technology in visitor monitoring systems and has the potential to serve as a more efficient solution than manual counting methods at Museum Buleleng.]]>
Copyrights © 2026
