π¨ AI-powered Hand Gesture Whiteboard β π Overview Welcome to the AI-powered Hand Gesture Whiteboard! This web-based application allows you to draw in the air using your index finger β no mouse or stylus required. It captures your hand movements in real-time and converts gestures into drawings on a digital canvas.
Built using cutting-edge technologies like TensorFlow.js and MediaPipe Hands, this app enables you to create drawings with the power of AI and hand gesture recognition.
- βοΈ Draw with your index finger β Control your drawing by simply moving your finger.
- π§½ Toggle between Drawing and Erasing modes β Easily switch between drawing and erasing.
- π₯ Real-time hand tracking β Tracks your hand movements through the webcam in real-time.
- π§ Smooth drawing motion β The app features smooth stroke rendering with easing effects.
- π‘ No installation required β Just open it in your browser and start drawing instantly!
- HTML, CSS, JavaScript: For the core functionality and layout.
- TensorFlow.js: For real-time hand gesture recognition and tracking.
- MediaPipe Hands: Provides hand landmark detection and tracking.
- Canvas API: Renders strokes and manages the drawing surface.
Hereβs a breakdown of the project structure:
π AI-powered-Hand-gesture-whiteboard/
βββ index.html # Main HTML layout
βββ styles.css # Simple UI styling
βββ script.js # Logic for drawing & gesture recognition using MediaPipe
πΈ How It Works
Launch the app: Open the app in your browser and allow webcam access.
Track hand movements: The app tracks your index finger tip (landmark 8) using MediaPipe.
Draw or erase: As you move your finger, it draws on the canvas. Switch to "Eraser" mode to clear the canvas.
Smooth experience: The drawing is smooth, thanks to real-time hand tracking and motion easing.
π§ͺ Run Locally
To run the app on your local machine, follow these steps:
Clone the repository:
bash
Copy
Edit
git clone https://github.com/CuriousCoder25/AI-powered-Hand-gesture-whiteboard.git
Navigate to the project directory:
bash
Copy
Edit
cd AI-powered-Hand-gesture-whiteboard
Open index.html in your browser (you can use a local server like Live Server in VS Code for a smoother experience).
Allow webcam access: Make sure your webcam is enabled and the browser prompts you for camera permissions.
β οΈ Prerequisites
Webcam for hand tracking.
A modern browser (Chrome, Firefox, etc.) with WebRTC support.
π Acknowledgements
This project utilizes the following technologies:
MediaPipe Hands for hand landmark detection.
TensorFlow.js for machine learning and gesture recognition.
JSDelivr CDN for serving external libraries.
πͺΆ License
This project is licensed under the MIT License. Feel free to fork, remix, or contribute to the code.
ποΈ Outro
A canvas of code, drawn by a free soul β CuriousCoder25. π¨
Now, go ahead, let your creativity flow!