Skip to content

Latest commit

 

History

History
73 lines (65 loc) · 4.27 KB

README.md

File metadata and controls

73 lines (65 loc) · 4.27 KB

Smart Home Manager

Abstract

Smart Home Manager can be described as Home Automation System (HAS) using Facial Recognition that identifies users and enhances their in-home experience by automating the home according to their requirements. The system basically collects the user information and builds a profile for the user to automate the home appliances by identifying the user by using his facial characteristics as input.

The system collects the information about the required item used by the family members in their home through a web interface and stores the data. The system creates a profile and also captures the face of the family member. The system then uses the facial characters as an input, identifies the family member and distinguishes them from intruders and guest by interacting with the family member. The family members and neighbors are intimated about the intruder, if detected. The security is ensured every time, where user authentication is provided by two ways. The captured face as input in case of a registered user and the family member’s response for the alert message in case of guest. The family members are free to change their requirement through the web interface. The family members are distinguished from each other through their facial characteristics and thus the system is flexible between registered user and a guest user. The web interface is designed in HTML, CSS, JavaScript. The data is stored in MongoDB using PHP-Mongo Driver. The facial recognition and all image processing works are carried out using OpenCV-Python library. The Bot is designed to interact with family members using Slack API. The Notification part is achieved using Pushbullet API.

Modules

1.Profile Creation:

  • Web interface creation
  • Collecting Data
  • Storing Data

2.Dataset Collection

  • Collect Name
  • Capture Faces

3.Dataset Training

  • Train using the Eigen classifier
  • Train using the LBPH classifier
  • Store as XML File

4.Facial Recognition

  • Motion sensor integration
  • Recognise Faces
  • Write names of recognised faces
  • Recognise unknown faces

5.Switching Appliances ON:

  • Fetch requirements of the user
  • Switch ON appliances connected to the ports
  • Switch OFF appliances connected to the ports

6. Intruder Detection :

  • Infrared sensor integration
  • Notify user
  • Bot message (Send and Receive message)

Specifications Used

Hardwares

  • Raspberry Pi 3 Model B
  • Raspberry Pi Camera Module
  • Motion Sensor
  • Infrared Sensor
  • 5V Relay (4 Channel)
  • Wires , bulb sockets, bulbs, GPIO Pins,Breadboard

Softwares

VNC Viewer
XAMPP

Technologies

Programming Language: Python, PHP
Web Server: Apache
Libraries: OpenCV, Numpy, PIL
Data Storage: MongoDB
API: Pushbullet, Slack

User interfaces

Login Screen:

image

Sign Up Screen:

image

Requirements Screen:

image

Requirements Screen (Edit Option):

image

Adding Person:

image

Modifying A Person:

image

Slack Bot (User Responding "Yes"):

image

Slack Bot (User Responding "No"):

image

Pushbullet Notification:

image