loading
A modular computer vision system mapping real-time webcam hand gestures to OS cursor control, clicks, scrolling, volume, and brightness adjustments.

Touchless interaction enables sterile computing environments, accessibility accommodations, and intuitive natural user interfaces. This project builds a low-latency gesture recognition pipeline running directly on standard consumer webcams.
Using Google's MediaPipe Hands, 21 3D hand landmarks are extracted per frame and fed into a gesture classifier. Coordinate smoothing algorithms eliminate jitter, translating finger distances and spatial geometry into precise mouse cursor movements, click triggers, multi-directional scrolling, and system volume/brightness scaling.