IoT Based Facial
Recognition System For
Access Control
By:
Md. Saddam Hossain
Roll No:1118030
Outlook
1
A short description about my project.
Project Overview
2
2
Biometric recognition system for access control
Recognition for Access Control
6
Describe about the present application and future scope
Application & Future Work
4
Describe the working procedure
How It Works
3
Describe Implementation steps
Methodology
5
Describe about the benefits & Limitation
Benefits & Limitation
7
questions answer
Conclusion
Project overview
 This is an IoT based project. Where I use facial Recognition
system for recognize person. Through the recognization, we can
control our home or office access through internet, from
anywhere of the world.
Objective:
The objectives of this project are:
 To develop a system or device that provide office and home
security.
 This device reduce human effort security maintenance cost.
 To develop a device that help blind or handicap people to
identify the new comer of home and open the door.
Recognition for Access Control
 For Access control, Recognition system is a new thinking. In
conventional system we use lock but now except that we can use
biometrics recognition system
 There are two kinds of biometric recognition system
Physiological
 Finger-scan
 Facial Recognition
 Iris-Scan
 Retina-Scan
 Hand Scan
Behavioral
 Voice-scan
 Signature-scan
 Keystroke-scan
Facial recognition system
A facial recognition system is a computer application capable
of identifying or verifying a person from a digital image or a video
frame from a video source.
Why I chose Facial recognition system:
a. It requires no physical interaction on behalf of the user.
b. It is accurate and allows for high enrolment and verification rates.
c. It does not require an expert to interpret the comparison result.
d. It can use your existing hardware infrastructure, existing cameras
and image capture Devices will work with no problems
e. It is the only biometric that allow you to perform passive
identification in a one to Many environments (e.g. identifying a
terrorist in a busy Airport terminal
Methodology
Initial Work:
 Install OPERATING SYSTEM
 INSTALL OPENCV ON REASPBERRY PI
 Setting camera to the Raspberry pi
Recognition System:
 DATASET GENERATOR
 TRAIN A FACE RECOGNIZER
 FACE RECOGNITION
Setting
Camera
Install
OpenCV
Install
OP
Face
Recognizer
Data
Trainer
Dataset
Generator
Control
servo &
send
Notification
How It Works
Working Steps of face recognition
Working Algorithm
1 1 1
1 1 1
1 1 1
1 2 3
2 4 9
3 6 9
I use viola jones algorithm for Face Detection and recognition
There are Four stages of this algorithm
(i) Haar feature selection
(There are 160 thousand feature)
(ii) An integral image
the value of pixel (x,y) is the sum of the pixel above to the left
(iii) Adaboost training
Relevant and irrelevant feature is identified
(iv) Cascading classifiers.
Reject the non face image
Figure:1 Figure:2
Figure:3
Stage 1
input a
face?
Definitely
Not
Discard
Input
May
Stage 2
input a
face?
Definitel
y Not
Discard
Input
May
Figure:4
Benefits and Limitation
Benefits
 Decreases security maintenance
cost and human effort.
 This system works automatically and
manually.
 We can placed this system
anywhere and control from all over
the world.
 The same approach can be used to
detect other objects (new training is
needed of course)
 In pattern recognition it is normally a
good idea to have smart features.
 Features are very simple but they
are very numerous, and the classifier
is very smart.
Limitation
 If anyone use any authorized people
picture the door will be open.
 In night or cloudy environment it
doesn’t work properly.
 My project accuracy for recognizing
a face is about 60%.
 The face detector can detect faces
that are tilted up tobout ± 15 degrees
in plane and about ± 45 degrees out
of plane (toward a profile view).
 Problems with harsh backlighting
and occlusions
Application And Future Work
Application:
• Home security
• Office security
• Handicap and blind people
• Shopping mall and restaurant
• Searching people and find terrorist
• Record attendance
Future Work:
• Increased accuracy
• threshold value to improve
• analyze the face in 3-D by using the combination of two cameras
• probability of error will be decreased
• system will be more accurate and with a very low cost
Conclusion:
Fig; Prototype of Smart Home with face recognition
Any Question
?
I am still working for developing
my project
Thanks

Facial Recognition System For Access Control

  • 1.
    IoT Based Facial RecognitionSystem For Access Control By: Md. Saddam Hossain Roll No:1118030
  • 2.
    Outlook 1 A short descriptionabout my project. Project Overview 2 2 Biometric recognition system for access control Recognition for Access Control 6 Describe about the present application and future scope Application & Future Work 4 Describe the working procedure How It Works 3 Describe Implementation steps Methodology 5 Describe about the benefits & Limitation Benefits & Limitation 7 questions answer Conclusion
  • 3.
    Project overview  Thisis an IoT based project. Where I use facial Recognition system for recognize person. Through the recognization, we can control our home or office access through internet, from anywhere of the world. Objective: The objectives of this project are:  To develop a system or device that provide office and home security.  This device reduce human effort security maintenance cost.  To develop a device that help blind or handicap people to identify the new comer of home and open the door.
  • 4.
    Recognition for AccessControl  For Access control, Recognition system is a new thinking. In conventional system we use lock but now except that we can use biometrics recognition system  There are two kinds of biometric recognition system Physiological  Finger-scan  Facial Recognition  Iris-Scan  Retina-Scan  Hand Scan Behavioral  Voice-scan  Signature-scan  Keystroke-scan
  • 5.
    Facial recognition system Afacial recognition system is a computer application capable of identifying or verifying a person from a digital image or a video frame from a video source. Why I chose Facial recognition system: a. It requires no physical interaction on behalf of the user. b. It is accurate and allows for high enrolment and verification rates. c. It does not require an expert to interpret the comparison result. d. It can use your existing hardware infrastructure, existing cameras and image capture Devices will work with no problems e. It is the only biometric that allow you to perform passive identification in a one to Many environments (e.g. identifying a terrorist in a busy Airport terminal
  • 6.
    Methodology Initial Work:  InstallOPERATING SYSTEM  INSTALL OPENCV ON REASPBERRY PI  Setting camera to the Raspberry pi Recognition System:  DATASET GENERATOR  TRAIN A FACE RECOGNIZER  FACE RECOGNITION Setting Camera Install OpenCV Install OP Face Recognizer Data Trainer Dataset Generator Control servo & send Notification
  • 7.
    How It Works WorkingSteps of face recognition
  • 8.
    Working Algorithm 1 11 1 1 1 1 1 1 1 2 3 2 4 9 3 6 9 I use viola jones algorithm for Face Detection and recognition There are Four stages of this algorithm (i) Haar feature selection (There are 160 thousand feature) (ii) An integral image the value of pixel (x,y) is the sum of the pixel above to the left (iii) Adaboost training Relevant and irrelevant feature is identified (iv) Cascading classifiers. Reject the non face image Figure:1 Figure:2 Figure:3 Stage 1 input a face? Definitely Not Discard Input May Stage 2 input a face? Definitel y Not Discard Input May Figure:4
  • 9.
    Benefits and Limitation Benefits Decreases security maintenance cost and human effort.  This system works automatically and manually.  We can placed this system anywhere and control from all over the world.  The same approach can be used to detect other objects (new training is needed of course)  In pattern recognition it is normally a good idea to have smart features.  Features are very simple but they are very numerous, and the classifier is very smart. Limitation  If anyone use any authorized people picture the door will be open.  In night or cloudy environment it doesn’t work properly.  My project accuracy for recognizing a face is about 60%.  The face detector can detect faces that are tilted up tobout ± 15 degrees in plane and about ± 45 degrees out of plane (toward a profile view).  Problems with harsh backlighting and occlusions
  • 10.
    Application And FutureWork Application: • Home security • Office security • Handicap and blind people • Shopping mall and restaurant • Searching people and find terrorist • Record attendance Future Work: • Increased accuracy • threshold value to improve • analyze the face in 3-D by using the combination of two cameras • probability of error will be decreased • system will be more accurate and with a very low cost
  • 11.
    Conclusion: Fig; Prototype ofSmart Home with face recognition Any Question ? I am still working for developing my project
  • 12.