Skip to content
Facial Recognition Relay
Machine Learning2018

Facial Recognition Relay

Summary

The application activates a GPIO relay on a Raspberry Pi when an enrolled user is recognized. Django provides the web UI on the Pi: users upload a photo from the camera (or file), and the image is forwarded to a Flask service that runs face_recognition (dlib encodings) with known faces stored in PostgreSQL. Authorized matches can drive the relay via relay.py.

Architecture

The repo splits edge control (django_front_end) from recognition (face_web_app). Django posts multipart images to Flask on localhost:5000 for enrollment (/add) and identification (/recognize).

Architecture diagram: Django on Raspberry Pi, Flask face recognition service, PostgreSQL, and GPIO relay

Components and enroll / recognize / relay flows from the GitHub source