Webcam-Driven Hand Tracking for Interactive Game

The evolution of CV and ML has opened a lot opportunities for natural human-computer interaction.Traditional input devices such as keyboards, mice, and controllers, though effective, can limit immersion and accessibility in interactive applications.Gesture recognition systems like Kinect and Leap Motion have advanced this field but remain constrained by their reliance on costly, specialized hardware.This paper presents a webcam-driven hand tracking system that enables real-time, gesture-based game control using only widely available hardware and open-source frameworks.Mediapipe is used to detect hand landmarks, OpenCV is used to process images, and Pygame is used to create an interactive 2D Fruit Ninja game.In this prototype, players pop balloons using a fist gesture while avoiding bombs that impose penalties, with hand movements directly mapped to in-game actions.Under standard conditions, the system demonstrated real-time responsiveness (25-30 FPS, latency 100 ms) and gesture recognition accuracy exceeding 90%, according to an experimental evaluation.Usability testing confirmed that players quickly adapted to the gesture controls, finding them intuitive and immersive.Compared with existing gesture-based systems, the proposed approach offers a low-cost, accessible, and engaging solution that highlights the scope of webcam-based interaction for entertainment, education, and many more technologies.