AI POWERED VIRTUAL MOUSE USING COMPUTER VISION

Abstract

Traditional computer interaction primarily relies on physical devices such as keyboards and mice. Although these devices have been reliable for decades, they create limitations in terms of ergonomics, hygiene, and accessibility. In the modern era of Human–Computer Interaction (HCI), there is a growing demand for touch less solutions that improve convenience and safety, particularly in the wake of global health concerns.

This project presents a vision-based virtual mouse system that allows users to control computer operations through hand gestures captured via a standard webcam. The system leverages OpenCV for image processing, Mediapipe for hand landmark detection, and PyAutoGUI for mouse event control. The index finger is mapped to cursor movement, while proximity between the index and thumb fingers is used to trigger click events.

Experimental results demonstrate that the proposed system operates in real time with high accuracy, achieving smooth cursor movement and reliable click detection. The solution is low-cost, requires no specialized hardware, and can serve as an accessible interface for everyday computing, interactive presentations, and assistive technologies.