/smart-door

Primary LanguageMakefile

Smart-Door-Lock

group name: KROM

Group members

Project name: SMART DOOR

Idea:

a smart door will open automatically if it recognizes the person standing in front of it. In the opposite case, if it doesn't, a notification is sent to the owner device who can choose whether to allow the person to enter the house or not.

Proposed Versions:

  • 1st-Version : The first version is a simple one : if a person wants to enter and is recognized, the door open automatically (in our case a led lights up green). If the person is foreign (a led lights up red), a notification is sent to the owner of the house who can accept or refuse to open the door (also, he has the possibility to add this person in the list of people that the device can recognize) the idea of this project resembles the project of august smart lock (But the challenge is to make it work with low cost equipment!). here is a useful link to explain: https://www.youtube.com/watch?v=WMln85LNENo link of our first version

  • 2nd-Version : the purpose of the second version is to improve the security part of the device. For example: Someone can put a picture of a person in front of the camera and this person is recognized by the door. So we have to detect a real person from pictured one.

  • 3rd-Version : The last version we can propose is still based on improving the security part( because a 100% secure device is the real challenge) So here's another example: A thief can force a person to stand in front of the door because that person can be recognized by the door. So the idea is to open the door based on a model of emotions recognition .

Framework:

  • ESP-WHO: is a face detection and recognition platform that is currently based on Espressif Systems' ESP32 chip
  • TensorFlow: is an open source machine learning framework.

Library :

  • Open-cv :an open Source Computer Vision Library specialized in real-time image processing.

List of essential components:

  • ESP32-Wrover: a powerful, generic WiFi-BT-BLE MCU module that targets a wide variety of applications, ranging from low-power sensor networks to the most demanding tasks, such as voice encoding, music streaming and MP3 decoding.
  • ov2640 camera: 2 Megapixels,supports image sizes: scaling down from SXGA to 40×30, image quality controls including [color saturation, gamma, sharpness, lens correction, white pixel canceling, noise canceling, and 50/60 Hz luminance detection]

Requirements

In the following software and hardware list, you can run the code file in this repository.

Software OS
Python 2.x/3.x, opencv, face_recognition, numpy, pandas, smtp, glob, email,lamp, esp-idf, arduino, Ubuntu 16.04 or greater
  • Use pip to install all packages of python:

E.g.

pip install opencv

you’ll need to make sure you have pip available. You can check this by running

pip --version

if you don't have pip, i suggest this link to install it

  • To install lamp i suggest this link
  • To install and config esp-idf follow steps in this link
  • To install arduino follow this

Test of our application:

  • config the web part (import data base in your phpmyadmin, lunch the webSite)
  • config the caméra :
    • config your wifi : using cmd -> make menuconfig
    • lunch the caméra : using cmd -> make monitor
    • check the website streaming if it work, to go to the next step
  • config system of face recognition :
    • installing all dependencies
    • lunch the script face.py
  • Now, if a person stands in front of the camera the system will detect it and do its work

Demonstration of our first version