2. INTRODUCTION
• Face recognition is a process of scanning a face and
matching it against a library of known faces.
• Facial features of an image are extracted and a template is
generated.
• While matching, the features of the input image is
extracted and a template is generated.
• The template is compared against all the templates in the
database.
3. • If the match is found then the user is acknowledged as a
valid user and the alarm circuit is deactivated.
• In enhancement of the project ie attendance management
system for employees, once the user is validated as an
authorised user he will be marked as present else absent.
• We are also generating an attendance report.
4. Problem Statement
• To build an efficient security system which uses face
recognition technique.
• Authenticating users based on their facial images
which are compared against those present in the
database.
• An alarm circuit that will be activated as soon as an
intruder tries to break through the system.
6. DESCRIPTION
• Sensor Module: It captures face images of individuals.
• Face Detection and feature extraction module:
Human faces are detected and their features are extracted
to form the template.
• Identification Module: The template extracted from
the input image is compared against stored templates in
the database.
• System Database Module: This module is responsible
for enrolling users in a face recognition system database.
9. STEPS IN PCA ALGORITHM
• Organize the data set and calculate empirical mean:
Data is organized into a set of matrices and empirical
mean(U) is calculated.
• Calculate the deviations from the mean: A new matrix
B stores the mean subtraction. This is used to minimize the
mean square error.
• Find the covariance matrix : Covariance matrix C is
computed by multiplying matrix B by its transpose.
10. • Find the eigenvectors and eigen values of the
covariance matrix: The diagonal matrix D is
calculated of the covariance matrix C.
• Rearrange the eigenvectors and eigen values : The
eigen vector matrix V and eigen value matrix D is
sorted in the order of decreasing eigen weights.
11. • Select a subset of the eigenvectors as basis vectors:
Save the first L columns of V as the M × L matrix W:
• Convert the source data to z-scores (optional):
Create an M × 1 standard deviation vector s from the
square root of each element along the main diagonal
of the diagonalized covariance matrix C.
• Project the z-scores of the data onto the new basis:
13. WORKING
• IC mct2e receives the output from the computer via
its First and Second pins.
• If the computer sends a high output i.e. +5Volts when
the face does not match the LED inside the IC glows.
• This illumination of the LED is detected by the
transistor which inturn sends the signal to the relay
and turns it into an ON state.
• When low signal is sent by the computer then the low
signal(0 Voltage) is sent by the computer to the IC
mct2e and the Relay is switched off.
15. WORKING
• When the Relay is on the current flows through the
timer and the alarm circuit is active.
• If anyone touches the touch plate the buzzer rings and
the intruder is detected.
• When the Relay is off the 555 timer is disconnected
from the power supply and the circuit is deactivated.
16. CONCLUSION
• Using Principle component analysis the face
recognition was implemented using c# with the help
of small set of images.
• The system performance depends on the images
stored in the database and the input image provided
by the user.
• This system identifies the users based on any facial
images.
• The monthly attendance report for the employees was
successfully generated.
17. FUTURE SCOPE
• Techniques like 3-D modelling, neural networks can be
used.
• Different circuits like door circuit, vibrator sensor circuit
can be integrated with face recognition to fulfill different
security objectives.