AI Virtual Mouse

AI Virtual Mouse is a Python application that enables controlling the mouse cursor with hand gestures captured via a webcam.
It leverages MediaPipe Hands for landmark detection, OpenCV for video processing, and PyAutoGUI for system-level mouse control.

Features

• Real-time hand detection and landmark tracking with MediaPipe
• Cursor movement mapped to index finger tip position
• Gesture-based controls:
  • Move Cursor → move index finger
  • Left Click → pinch index finger & thumb together
  • Right Click → pinch middle finger & thumb
  • Scroll Up/Down → vertical hand gestures
• Multi-hand support (detects left vs right hand)
• Configurable gesture mapping for customization

Requirements

• Python 3.10+
• OpenCV
• MediaPipe
• PyAutoGUI
• NumPy

Install them with:

pip install opencv-python mediapipe pyautogui numpy

Usage

1. Clone the repository:

   git clone https://github.com/yourusername/ai-virtual-mouse.git
   cd ai-virtual-mouse

2. Run the application:

   python main.py

3. Place your hand in view of the webcam:

   • Move your index finger → cursor moves
   • Pinch index & thumb → left click
   • Pinch middle & thumb → right click
   • Move hand up/down with open palm → scroll

4. Press q to quit.

Configuration

• Sensitivity, smoothing, and gesture mappings can be adjusted in config.py.
• Debug overlays (landmarks, FPS, gesture labels) can be enabled by setting draw_landmarks=True when initializing HandTracker.

Project Structure

ai_virtual_mouse/
├── main.py                # Entry point
├── hand_tracker.py        # MediaPipe hand detection wrapper
├── gesture_classifier.py  # Maps landmarks → gestures
├── mouse_controller.py    # Handles PyAutoGUI mouse actions
├── utils/                 # Helper functions
└── tests/                 # Unit & E2E tests

Limitations

• Requires consistent lighting and clear webcam view of hands.
• Latency may increase on low-end hardware.
• Currently optimized for single-user interaction.

Contributions

Contributions are welcome!

• Fork the repo
• Create a feature branch
• Submit a pull request

---

Built using Python, MediaPipe, OpenCV, and PyAutoGUI.