부산대학교 도서관

Using a physical mouse is a popular way to interact with a computer but is limited by the reliance on external devices such as batteries and dongles. In this paper, an artificial intelligence based virtual mouse system that enables device-free computer control using computer vision and machine learning technology has been proposed. The system detects hand gestures and hand gestures through a built-in webcam or camera and uses machine learning algorithms to interpret these gestures into computer actions such as clicking left, clicking right, scrolling up-down, and movement of cursor. This algorithm is based on deep learning techniques for hand recognition, eliminating the reliance on human intervention and physical devices. This system can play a significant role in preventing the spread of COVID-19 and its variant by reducing contact with foreign devices at Airport. To evaluate the effectiveness of the system, we conducted an experiment that compared an artificial intelligence virtual mouse system with a physical mouse. The results proposed system achieves accuracy and speed comparable to physical mice. Additionally, user feedback indicates that the AI virtual mouse system is intuitive and easy to use. In conclusion, the proposed artificial intelligence virtual mouse system provides a new and convenient solution for device-free computer control and is useful in applications such as medical and public spaces where minimizing contact with external devices is important. It may be used in a variety of settings.