Virtual Mouse Using Open CV

Author Name : Abhijay Tyagi, Muskaan Nagpal

ABSTRACT

The Virtual Mouse project using OpenCV is a practical solution that employs computer vision technology to create a virtual mouse that can be used without the need for a physical mouse or any other pointing device. The project aims to develop a virtual mouse system that is accurate, efficient, and user-friendly, and can provide an alternative to a physical mouse that can be uncomfortable to use for extended periods. The virtual mouse system developed in this project comprises a webcam, a computer, and a software program developed using Python and OpenCV. The webcam captures the video feed of the user's hand movements, which is then processed by the software program. The program uses computer vision algorithms to recognize hand gestures and movements and maps them to corresponding mouse functions. The development of the virtual mouse system involves several steps, including hand detection, hand tracking, and gesture recognition. The system uses skin color segmentation to identify and isolate the hand region from the video feed in the hand detection step. In the hand tracking step, the system tracks the movement of the hand, determines its position and velocity, and uses this information to control the mouse pointer. Finally, in the gesture recognition step, the program identifies and recognizes various hand gestures, such as clicking, scrolling, and dragging, and maps them to corresponding mouse functions. The project includes extensive testing under different conditions, including various lighting conditions and distances from the screen, to ensure the accuracy and reliability of the virtual mouse system. The results show that the system is capable of accurate and efficient gesture recognition and can perform mouse functions with minimal delay. The project demonstrates the potential of computer vision technology to create innovative and practical solutions for human-computer interaction. The project shows that a virtual mouse system using OpenCV can be an effective alternative to a physical mouse, providing a more intuitive and natural way of interacting with a computer.