posture-detection-app 🧍♂️ Posture Detection App A full-stack real-time and video-based posture detection system using React, Flask, MediaPipe, and OpenCV. It detects posture issues like:
🚫 Knee over toe
📸 Features ✅ Real-time webcam posture capture.
🎥 Upload video for full session analysis.
📊 Instant feedback (Good/Bad posture).
🔁 Rule-based logic using MediaPipe landmarks.
📂 Project Structure
posture-detection-app/ ├── frontend/ # React frontend ├── backend/ # Flask API │ ├── app.py │ └── posture_analysis.py └── README.md
🛠️ Setup Instructions ⚙️ Backend (Flask) Navigate to backend:
cd backend python -m venv venv venv\Scripts\activate # On Windows pip install -r requirements.txt python app.py
🌐 Frontend (React) Navigate to frontend:
cd frontend npm install npm start
📡API Endpoints: POST /posture – Analyze a single webcam frame.
POST /analyze – Analyze a full video file.
🔍 Tech Stack: Frontend: React, react-webcam, Axios.
Backend: Flask, Flask-CORS, OpenCV, MediaPipe.
Video Analysis: MediaPipe Pose, NumPy, math.
📌 Deployment Frontend: Vercel / Netlify / Render
Backend: Render / Railway / Replit
🧠 Future Improvements: Real-time video streaming support.
Pose history & analytics.
User accounts for tracking progress.