/robt310

Face Recognition Access System

Primary LanguageTypeScript

Integrating Facial Recognition with IoT for Access Control: Unlocking Doors using Electromagnets and Logging Worker Entries

Prerequisites

  • Node.js
  • Python 3.6+

Hardware

There are two ways to run this project:

  1. Using a Raspberry Pi 4 + Webcam
  2. Using a computer with a webcam + ESP8266

Choose one of the above and follow the instructions below.

1. Raspberry Pi 4

Connect relay to GPIO 7. Connect webcam to Raspberry Pi.

2. ESP8266

Connect relay to GPIO 5 (D1). Edit wifi configuration in /relay/main.ino and upload to ESP8266.

Installation

  1. Clone this repository:
git clone git@github.com:superhooman/robt310.git
cd robt310
  1. Install the required packages for each subsystem:
  • /cron: 
    cd cron
npm ci
    Also, if you are using ESP8266, then create a .env file in the cron directory and add the following: 
    RELAY_ADDRESS=<address of the relay> # e.g. http://192.168.0.193/
  • /deepface: 
    cd deepface
pip3 install deepface # or pip install deepface
python3 check.py # to download the required models
  • /system: 
    cd system
npm ci
npx prisma db push
  1. Now, we need to setup processes to run each subsystem. We will use pm2 to manage these processes. Install pm2 globally:
npm install -g pm2
  1. Start the processes:
  • /cron: 
    cd cron
pm2 start index.js --name "cron"
  • /system: 
    cd system
npm run build
pm2 start npm --name "system" -- start
  1. You all set! Now you can access the system at http://localhost:3000 and the cron job will check camera every 10 seconds.