How avinashhsinghh/heart-rate-detection works
Project type and purpose
• This is a full-stack, real-time, browser-based application (“Pulse”) for non-contact heart-rate detection.
• Frontend: HTML5/JavaScript (Canvas, WebRTC, headtrackr.js, D3.js, Rickshaw).
• Backend: Python (Flask + flask_sockets), NumPy, Independent Component Analysis (JADE), Fast Fourier Transform (FFT).
File structure
• Top-level:
• app.py – Flask entry point and WebSocket routes
• jade.py & model.py – ICA and FFT algorithms
• deploy.sh, Procfile, runtime.txt – deployment scripts/config for Heroku or similar
• requirements.txt – Python dependencies
• static/ – JS, CSS, images
• templates/ – Jinja2 HTML templates
• Patterns:
• Monolithic backend serving a single WebSocket API
• Frontend signal-processing pipeline in JS, offloading heavy math to Python server
README
• Documents data flow: webcam → Canvas → headtrackr ROI → channel averages → WebSocket → Python (normalize → ICA → FFT → select heartbeat → send back) → browser (compute BPM → render graphs).
• Lists major libraries and frameworks used in each tier.