Skip to content


Repository files navigation

Intelligent Facial Recognition System

GitHub release Language python Language python Build Status License

This is a project of the facial recognition with Movidius on RaspberryPi 3B+ platform. It also uses Django and Django REST framework which providing the web platform. The project would like to build a safety and intelligent face recognition system in AI era.

If you appreciate the content 📖, support projects visibility, give 👍| ⭐| 👏


The code is tested using Tensorflow r1.7 and Movidius NCSDK2 under Debin 2018-06-27(Kernel version:4.14) with django 2.1.5 and Python 3.5 & 3.6.

File architecture

django_venv/  # django virtual env for RPI
face_recognition_model/  # Movidius NCS code     
iot_control/  # iot control code
ran-django-template/  # django platform using my own model
requirements.txt  # requirements for run the code

How to run it

  1. Install requirements
pip install -r requirements.txt
  1. Install Neural Compute Application Zoo

    Method attached below.

  2. Install rpi-mjpg-streamer

    Method attached below.

  3. Run ran-django-template

cd ran-django-template
python runserver
  1. Run face_recognition_model
cd face_recognition_model
make run
  1. Run iot_control
cd iot_control

Admin Account

python createsuperuser

username: ranxiaolang
email: YOUR EMAIL  
password: ranxiaolang  

Access the web page though this link: .

Real Product Images

image image


  • Logitech HD Webcam C270
  • Micro SD Card 32G
  • Raspberry Pi 3 B+
  • Intel Movidius Neural Compute Stick

The code requires Python 3.5 or Python 3.6, Tensorflow 1.7 or later, as well as the following python libraries:

  • Pillow
  • django
  • django-allauth 0.37.1
  • django-widget-tweaks 1.4.3
  • pip 18.0
  • qrcode 6.0
  • setuptools 40.4.3
  • djangorestframework
  • markdown
  • django-filter

Those modules can be installed using: pip3 install xxx or pip install xxx .

Follow How to run it

Neural Compute Application Zoo

This repository is a place for any interested developers to share their projects (code and Neural Network content) that make use of the Intel® Movidius™ Neural Compute Stick (Intel® Movidius™ NCS) and associated Intel® Movidius™ Neural Compute Software Development Kit.

You can use the following url(NC App Zoo) or git command to use the ncsdk2 branch of the NC App Zoo repo:

git clone -b ncsdk2

Install Django and Django REST framework

pip3 -V

sudo pip3 install -U setuptools

sudo pip3 install -U django

sudo pip3 install -U djangorestframework

sudo pip3 install -U django-filter

sudo pip3 install -U markdown

sudo pip3 install -U requests

Install Adafruit_Python_DHT library

git clone

cd Adafruit_Python_DHT

sudo python3 install


Install Adafruit_Python_BMP library

git clone

cd Adafruit_Python_BMP

sudo python3 install


Install psutil (process and system utilities)

sudo pip3 install psutil

Install rpi-mjpg-streamer

Instructions and helper scripts for running mjpg-streamer on Raspberry Pi.

A. Setup mjpg-streamer

Enable Raspberry Pi Camera module from raspi-config

$ sudo raspi-config

Install necessary packages for mjpg-streamer

$ sudo apt-get update
$ sudo apt-get install build-essential libjpeg8-dev imagemagick libv4l-dev git cmake uvcdynctrl

Build mjpg-streamer

$ sudo ln -s /usr/include/linux/videodev2.h /usr/include/linux/videodev.h
$ git clone
$ cd mjpg-streamer/mjpg-streamer-experimental
$ make install

B. Run mjpg-streamer

1. Clone this repository

$ git clone

2-a. Run mjpg-streamer from the shell directly

# copy & edit to your environment or needs
$ cp rpi-mjpg-streamer/ somewhere/
$ vi somewhere/

# then run
$ somewhere/

2-b. Or run mjpg-streamer as a service


# copy & edit systemd/mjpg-streamer.service file,
$ sudo cp rpi-mjpg-streamer/systemd/mjpg-streamer.service.sample /lib/systemd/system/mjpg-streamer.service
$ sudo vi /lib/systemd/system/mjpg-streamer.service

# then register as a service
$ sudo systemctl enable mjpg-streamer.service

# or remove it
$ sudo systemctl disable mjpg-streamer.service

# and start/stop it
$ sudo systemctl start mjpg-streamer.service
$ sudo systemctl stop mjpg-streamer.service

C. Connect

Connect through the web browser:


Most modern browsers(including mobile browsers like Safari and Chrome) will show the live stream immediately.



Method 1

pip3 install virtualenv   

Copy django_venv and activate

source venv/bin/activate    

Method 2

pip install -r requirements.txt

sqlite3 数据库文件db.sqlite3 权限 666

chmod 666 db.sqlite3

django 所在文件夹 权限 775

chmod 777 xxx


Just can be used for non-business projects. If you use ran-django-template, please give me a star. Thanks!