A real-time computer vision application demonstrating hand gesture recognition using OpenCV and MediaPipe for game control.
- Real-time hand detection using MediaPipe
- Gesture classification (5 distinct hand poses)
- Finger curl detection for firing mechanism
- Smooth position tracking with LERP interpolation
- Direction calculation from hand movement
| Gesture | Action | CV Technique |
|---|---|---|
| ✊ Fist | Burst attack | Finger extension detection |
| 🖐️ Open hand | Shield | Palm openness + finger spacing |
| ✌️ Peace | Multi-shot | Two-finger isolation |
| 👌 OK sign | Power attack | Thumb-index distance measurement |
| Index curl | Fire | Joint angle calculation |
pip install -r requirements.txt
python demo.py- Camera capture → OpenCV VideoCapture
- Hand detection → MediaPipe Hands
- Landmark extraction → 21-point hand model
- Gesture classification → Custom algorithm
- Game control → Real-time input mapping
Built for: Computer Vision portfolio demonstration
Development time: ~4 hours
Focus: Real-time gesture recognition accuracy
Author: Darren Chai Xin Lun
Email: ddcxl0301@gmail.com
GitHub: @darrencxl0301