HandDetecting

Real-time hand gesture control system built with MediaPipe and OpenCV.

This project uses your webcam to detect hand landmarks and lets you control:

• System volume (right hand)
• Screen brightness (left hand)
• Mouse movement, click, and scrolling (right hand in mouse mode)

Features

• Dual-hand tracking with MediaPipe Tasks API (hand_landmarker.task)
• Auto-download of the hand landmark model if missing
• Two control modes:
  • Media mode: volume, brightness, mute toggle
  • Mouse mode: virtual touchpad, click, and scroll
• Real-time HUD overlay with FPS, active mode, and controls guide
• Signal smoothing for stable volume/brightness and mouse movement

Requirements

• Windows 10/11
• Python 3.9+
• Webcam

Install Dependencies

Run this in the project folder:

pip install opencv-python mediapipe numpy pycaw comtypes pyautogui screen-brightness-control

Notes:

• screen-brightness-control is optional. If not installed, brightness control is simulated and a warning appears on screen.
• First run may download hand_landmarker.task automatically from Google MediaPipe models.

Run

python HandTracking.py

Controls

Global

• Press M to toggle between Media and Mouse modes
• Press Q to quit

Media Mode

• Right hand pinch (thumb + index): adjust system volume
• Hold right-hand pinch near minimum for a short time: toggle mute/unmute
• Left hand pinch (thumb + index): adjust screen brightness

Mouse Mode (Right Hand)

• Index finger up only: move mouse cursor
• Index + middle fingers up and close together: left click
• Index + middle fingers up and apart: vertical hand movement scrolls page

How It Works

1. Captures webcam frames using OpenCV.
2. Runs MediaPipe Hand Landmarker on each frame.
3. Classifies left/right hand and extracts landmarks.
4. Maps gestures to OS actions:
   • Pycaw for master volume and mute
   • screen-brightness-control for brightness
   • PyAutoGUI for cursor, click, and scroll
5. Draws a live HUD and hand skeleton overlay.