Fullstack-assignment

Hand Gesture Controlled Bike Racing Game

A machine learning project that uses computer vision to track hand gestures and control a bike racing game in real-time.

Features

• Real-time hand gesture recognition using MediaPipe and OpenCV
• Bike racing game with physics and obstacles
• Gesture controls:
  • ✋ Open palm (5 fingers): Accelerate
  • ✊ Fist: Brake
  • ☝️ Index finger: Turn left
  • 🖐️ Pinky finger: Turn right

Project Structure

hand_gesture_bike_racing/
├── main.py                 # Main game loop
├── gesture_detector.py     # Hand gesture recognition
├── bike_game.py           # Bike racing game logic
├── camera_interface.py    # Camera handling and processing
├── models/                # Trained ML models
├── assets/                # Game assets (images, sounds)
├── data/                  # Training data and datasets
└── utils/                 # Utility functions

Installation

1. Install Python 3.8+
2. Install dependencies:

   pip install -r requirements.txt
   

Usage

1. Run the main game:

   python main.py
   

2. Position your hand in front of the camera
3. Use gestures to control the bike:
   • Show all 5 fingers to accelerate
   • Make a fist to brake
   • Point with index finger to turn left
   • Point with pinky finger to turn right

Technical Details

• Computer Vision: OpenCV for camera input and image processing
• Hand Tracking: MediaPipe for real-time hand landmark detection
• Machine Learning: TensorFlow for gesture classification
• Game Engine: Pygame for 2D graphics and physics
• Real-time Processing: Optimized for 30+ FPS

Requirements

• Webcam or camera device
• Python 3.8+
• 4GB+ RAM
• GPU recommended for better performance

Contributing

Feel free to contribute by improving gesture recognition, adding new game features, or optimizing performance!