Virtual Mouse System using Hand Gesture recognition with Open CV

Abstract

The use of hand gesture recognition for controlling virtual devices has gained popularity with the advancement of artificial intelligence technology. In this paper, we propose a hand gesture-controlled virtual mouse system that employs AI algorithms to identify and interpret hand gestures, translating them into mouse movements. The primary objective of this system is to provide an alternative interface, particularly beneficial for individuals facing challenges with traditional mouse or keyboard usage.Our proposed system utilizes a camera to capture images of the user's hand. To enable gesture recognition, the system undergoes training using a dataset comprising various hand gestures. Once a gesture is identified, it is translated into a corresponding mouse movement, subsequently executed on the virtual screen. An essential feature of our system is its scalability and adaptability, allowing seamless integration into diverse environments and devices. To enhance user interaction, our system incorporates dynamic/static hand gestures along with a voice assistant for virtually controlling all input operations. Notably, our approach employs Machine Learning (ML) and Computer Vision algorithms to recognize hand gestures and voice commands, eliminating the need for additional hardware requirements. The model is implemented using the MediaPipe framework, demonstrating the feasibility of our proposed system. In conclusion, our hand gesture-controlled virtual mouse system presents a promising avenue for improving user experience and accessibility through human-computer interaction. By leveraging AI, ML, and Computer Vision, we aim to create an intuitive and versatile interface that can be easily adapted to different environments, ultimately enhancing accessibility for a wider range of users. Keywords: Computer vision, hand gesture recognition, Media-pipe, virtual mouse.