Documentation

Published on November 2016 | Categories: Documents | Downloads: 75 | Comments: 0 | Views: 593
of 59
Download PDF   Embed   Report

face recognition,dip,image processing,document,java program,javacode,FRS

Comments

Content


AFRICA NAZARENE UNIVERSITY
CIT DEPARTMENT
Facial Recognition Access Control in Automated Information
System
BY
MacDonald Daniel Ngowi
09SBT118
Supervisor
Mr. Ombui !d"ard
27
th
July 2012
Submitted in partial fulfillment of the requirements of the Bachelor of Business
Information Technology
#!C$ARA%IO&
This is my original work and has not been presented in any institution or elsewhere.
No part of the work contained herein can therefore be reproduced without my permission.

Student &ame #A%!
#eclaration by t'e Supervisor
This proposal has been submitted by my approval as University Supervisor.
((((((((((.. (((((((((.
2
Supervisor #A%!
AC)&O*$!#+!M!&%
irst and foremost! I acknowledge the "lmighty #od for all he has enabled me to do as pertains
this pro$ect.
Secondly! I thank the University for its Support and the resources it has offered to
facilitate this pro$ect.
I also appreciate my lecturer and supervisor %r. &dward 'mbui for the guidance! mental
and physical support he has shown to ensure the adequate and successful completion of this
pro$ect.
I cannot forget my parents! relatives! and my guardian for the prayers! financial and
moral support they have offered towards undertaking this pro$ect.
3
#!FI&I%IO& OF %!RMS
Algorit'm( It is a step by step approach of defining the means of solving a problem. It usually
means a small procedure that solves a recurrent problem.
Auto,Aut'entication( It is a process of continuous validation of user authenticity while
operating an information system. The user operating the system must match the user logged into
the system in order to prevent session intrusion.
Face detection( is a computer technology that determines the locations and si)es of human faces
in arbitrary *digital+ images. It detects facial features and ignores anything else! such as
buildings! trees and bodies.
Face recognition( It is a computer technology of automatically identifying or verifying a person
from a digital image or a video frame from a video source. 'ne of the ways to do this is by
comparing selected facial features from the image and a facial database.
Motion trac-ing( It is a technology used to detect and track moving ob$ects through a sequence
of images. aces need to be tracked as normally the human head is not stationary.
4
ABS%RAC%
,ith the growing need to e-change information and share resources! information security has
become more important than ever in both the public and private sectors. "lthough many
technologies have been developed to control access to files or resources! to enforce security
policies! and to audit network usages! there does not e-ist a technology that can verify that the
user who is using the system is the same person who logged in. acial recognition technology
has been adopted to monitor the presence of the authenticated user throughout a session.
Therefore! only the legitimate user could operate the computer and unauthori)ed entities have
less chance to hi$ack the session. The goal is to enhance the level of system security by
periodically checking the user.s identity without disrupting the user.s activities.
5
%AB$! OF CO&%!&%S
DECLARATlON........................................................................................................................... 2
ACKNOWLEDGEMENT.............................................................................................................. 3
DEFlNlTlON OF TERMS............................................................................................................. 4
ABSTRACT.................................................................................................................................. 5
TABLE OF CONTENTS............................................................................................................... 6
LlST OF FlGURES...................................................................................................................... 7
LlST OF TABLES......................................................................................................................... 7
LlST OF ABBREVlATlONS.......................................................................................................... 8
CHAPTER l: lNTRODUCTlON................................................................................................... 9
CHAPTER 2: LlTERATURE REVlEW........................................................................................ l6
Source: Laurenz W.,Jean. F. Norbert Krüger, & Christoph M. (2003) Face recognition by elastic
bunch graph matching. Retrieved online on 22nd February 20l2 from:.....................................22
CHAPTER 3: METHODOLOGY................................................................................................. 24
CHAPTER 4: SYSTEM ANALYSlS AND DESlGN.....................................................................29
CHAPTER 5: WORK PLAN AND BUDGET...............................................................................42
CONCLUSlON........................................................................................................................... 48
REFERENCES.......................................................................................................................... 49
lT News Africa. (20l0) Kenya could implement face recognition technology for ATMs. Retrieved
Online on 28th January 20l2 from: <http://www.itnewsafrica.com/20l0/05/kenya-could-
implement-face-recognition-technology-for-atms/>....................................................................49
Juwei Lu, (2002) Boosting Linear Discriminant Analysis for Facial Recognition.........................49
Laurenz W.,Jean. F. Norbert Krüger, & Christoph M. (2003) Face recognition by elastic bunch
graph matching. Retrieved online on 22nd February 20l2 from:................................................49
BlBLlOGRAPHY........................................................................................................................ 50
X. D. Yang, P. Kort, and R. Dosselmann, "Automatically log off upon disappearance of facial
image," Contract Report CR 2005-05l, DRDC, Ottawa, Canada, March 2005..........................5l
APPENDlX l: GANTT CHART................................................................................................... 5l
6
APPENDlX ll: SAMPLE QUESTlONS....................................................................................... 52
APPENDlX lll: SAMPLE CODES............................................................................................... 53
APPENDlX lV: SAMPLE SCREENSHOTS................................................................................56
$IS% OF FI+.R!S
Figure l Standard Eigenfaces: Feature vectors are derived using eigenfaces...........................l9
Figure 2 Example of Six Classes using LDA..............................................................................20
Figure 3 Elastic Bunch Graph Matching.....................................................................................22
$IS% OF %AB$!S
Table l objective l: to develop user requirements/Preliminary investigation.............................43
Table 2 objective 2: to design the system and database requirements......................................44
Table 3 objective 3: to implement system functionalities............................................................45
Table 4 objective 4: to carry out testing and maintenance.........................................................45
Table 5 objective 5: to finalize documentation and presentation................................................46
Table 6 the cost estimation of various resources to be used in the research project..................47
7
$IS% OF ABBR!/IA%IO&S
&B#%( &lastic Bunch #raph %atching
&/&T( ace /ecognition Technology
/0T( ace /ecognition 0endor Tests
/S( acial /ecognition System
1%%( 1idden %arkov %odel
IS( Information System
2"( 2ocal eature "nalysis
233( 2ocality 3reserving 3ro$ection
24"( 2inear 4iscriminant "nalysis
35"( 3rincipal 5omponent "nalysis
8
S425( System 4evelopment 2ife 5ycle
C0A1%!R 23 I&%RO#.C%IO&
This chapter gives an overview of the field of computer vision and biometrics with focus on
facial *or rather face+ recognition technology. It gives out the meaning of face recognition!
applications of the technology! purpose and ob$ectives of the study! various questions pertaining
the study! limitations and delimitations of the technology! theoretical as well as conceptual
frameworks for the study.
2.2 I&%RO#.C%IO&
" smart environment is one that is able to identify people! interpret their actions! and react
appropriately. Thus! one of the most important building blocks of smart environments is a person
identification system. ace recognition devices are ideal for such systems! since they have
recently become fast! cheap! unobtrusive! and! when combined with voice6recognition! are very
robust against changes in the environment. %oreover! since humans primarily recogni)e each
other by their faces and voices! they feel comfortable interacting with an environment that does
the same.
9
acial recognition systems are built on computer programs that analy)e images of human
faces for the purpose of identifying them. The programs take a facial image! measure
characteristics such as the distance between the eyes! the length of the nose! and the angle of the
$aw! and create a unique file called a 7template7. Using templates! the software then compares
that image with another image and produces a score that measures how similar the images are to
each other. Typical sources of images for use in facial recognition include video camera signals
and pre6e-isting photos such as those in national identification databases.
ace recognition systems are computer6based security systems that are able to
automatically detect and identify human faces. These systems depend on a recognition algorithm!
such as eigenface or the fisherface. The first step for a facial recognition system is to recognize a
human face and extract it from the rest of the scene. Ne-t! the system measures nodal points on
the face! such as the distance between the eyes! the shape of the cheekbones and other
distinguishable features.
The nodal points are then compared to the nodal points computed from a database of
pictures in order to find a match. 'bviously! such a system is limited based on the angle of the
face captured and the lighting conditions present. New technologies are currently in development
to create three6dimensional models of a person8s face based on a digital photograph in order to
create more nodal points for comparison. 1owever! such technology is inherently susceptible to
error given that the computer is e-trapolating a three6dimensional model from a two6dimensional
photograph.
%ost algorithms currently in use are not showing high degree of performance and
accuracy in face detection and recognition especially in uncontrolled conditions! for instance!
changes in pose! facial e-pressions! variations in lightning! and among other variations. 94
l0
morphable model! neural networks! and laplacianfaces algorithms have shown some
improvements in such scenarios. 1owever! these approaches are still weak in differentiating a set
of twins.
ace recognition systems are deployed in large6scale citi)en identification applications!
surveillance applications! law enforcement applications! booking stations! kiosks! management
information system access! and even for operating system login needs. 1owever! most of these
systems lack auto6authentication to check the authenticity of the user currently operating the
system for the purpose of preventing unauthori)ed access and violation of rights and privacy to
data and information.
2.4 1ROB$!M S%A%!M!&%
0arious security technologies have been developed! such as authentication! authori)ation! and
auditing. 1owever! once a user logs on! it is assumed that the system would be controlled by the
same person. To address this flaw! I intend to develop a system that uses facial recognition
technology to periodically verify the identity of the user. If the authenticated user8s face
disappears! the system automatically performs a log6off or screen6lock operation.
2.5 1.R1OS! A&# OB6!C%I/!S
The purpose of this research study is to develop a facial recognition system that will keep track
of the authenticity of the user of a particular computer system or particular information system
and to improve the accuracy of recognition especially in uncontrolled conditions involving
variations in lightning! facial e-pressions! facial angles! and pose by implementing a
2aplacianface approach with 2ocality 3reserving 3ro$ections *233+ to facial recognition.
The ob$ectives of the system based on functional requirements include the following(
ll
i. To enable users to enroll their faces into the system( The system! after
implementation! will be capable of enrolling face templates through camera video
stream in form of photo snapshots in different face angles. &ach face template will
contain an e-traction of facial parts such as nose! mouth! eyes! and chicks from the
test sub$ect. This will be used in detection testing as well as recognition testing of
sample templates with different face test sub$ects.
ii. To provide face detection and motion tracking( "lso the system! after
implementation! will be capable of differentiating a face from the rest of the body by
detecting only facial parts which will later be used in recognition. This will be done
through camera video stream hence the face will be detected while in motion using
motion tracking module.
iii. To enable users to be authenticated through face recognition and two6factor
authentication( The system will also be capable of matching or compare the face
templates with face test sub$ects and will give confidence level to the recogni)ed
faces as well as re$ect those faces not in the face templates database. This will hence
give either positive or negative match to authenticate a user. The use of traditional
means can also accompany facial recognition for improved security.
iv. To enable auto recognition( The system will be capable of verifying the user currently
using the system. If the user is not the same user who logged in to the particular user
account! the system will take snapshots of the perpetrator and logs off the system
automatically. If the perpetrator is not skilled enough! he will not be aware of this
system running as a background process. 1ence will not force the system to close.
&ven if the perpetrator gains access to the user account without using the facial
l2
recognition module! this system will continue to monitor the current user of the
system as it will use the information of a particular user account.
v. To improve the accuracy of face recognition( The accuracy of face detection and face
recognition will be improved by various face angles enrollment! and in non6e-treme
usage scenarios i.e. enough lightning and normal pose. The use of effective
algorithms like the 2aplacianfaces with 2ocality 3reserving 3ro$ection *233+ will
help to improve the accuracy and the confidence level.
2.7 SI+&IFICA&C! OF %0! 1RO6!C%
/esearch in the various biometrics has been there since :;<=s. %a$or breakthroughs in the field
came in the :;;=s after the introduction of powerful processors and improvement in computer
memory. This hence led to wide applications of biometrics i.e. fingerprint! voice! face! among
others especially in buildings. security and surveillance.
This research on facial recognition for automated access control brings in new
opportunities in the field of biometrics as I will attempt to solve the problem of lack of
continuous authentication of the user of a particular information system in which at the end will
improve security and privacy of users.
This research also provides an alternative to the traditional means of access control i.e.
what you know! e-amples( passwords! emails! 3IN> and what you have! e-ample( Smart cards!
tokens> by introducing face recognition as the means of access control which is unique to a
person.s physical appearance. ,hen accessing our email accounts! social networking sites!
membership sites! online payment processor sites! desktop information systems! "T% machines
l3
or as website administrators! we will often use passwords! emails! smartcards! and 3IN for
authentication. Such traditional means can be stolen! fabricated! lost! forgotten! and duplicated.
2.8 $OCA$I%Y OF %0! 1RO6!C%9 B!&!FICIARY %O %0! 1RO6!C%
If this pro$ect is successful! any person who has access to a computer that has a camera is an
intended user of the system. This system will work well in laptops where most of the modern
ones come with an in6built webcam. It will work e-cellent in high pi-els cameras with at least ?6
megapi-els. " desktop camera is also efficient to secure desktop computers with this system.
Since this pro$ect aims at developing a customi)ed system that can be embedded in any
information system! the main people who will benefit from the system can be divided into three(
!ntertainment !nd,.sers
This system will become useful for anyone requiring innovative and futuristic solution of
authentication apart from the traditional means of passwords and emails or usernames. Such
users will be able to use their faces to gain access to their systems because it is a fun way to do it.
Such end6users are mostly home and school users who do not care much about their privacy
when their computers are concerned. They mostly need convenience of gaining access to their
systems.
l4
1o"er91rivacy !nd,.sers
These users will benefit or will find the system useful as an improved security system as
compared to the traditional password system. These users want more control over their private
information and data. These are users who are highly concerned of who gain access to their
systems. These are users who want to catch perpetrators of their systems. Such users can be
found at workstations *workplaces+! military bases! research and development institutions! and
among other areas that require confidentiality or discretion of personal or organi)ation
information.
System Administrators
If the system is to be deployed as a means of authenticating users in an information system for an
organi)ation! there has to be an administrator to manage the various user accounts i.e. create
facial based accounts for users! delete accounts! update accounts! and among other activities. The
administrator will have the highest access level as compared to other users. This will give the
administrator the power to perform a variety of activities as stated earlier.
l5
C0A1%!R 43 $I%!RA%.R! R!/I!*
4.2 BAC)+RO.&#
"utomated face recognition is a relatively new concept. 4eveloped in the :;@=s! the first semi6
automated system for face recognition required the administrator to locate features *such as eyes!
ears! nose! and mouth+ on the photographs before it calculated distances and ratios to common
reference point! which were then compared to reference data.
/ecently! researchers have developed many methods for face recognition. In order to
facilitate their comparison! the &/&T program was established. The &/&T protocol consists
of a database of facial images of more than :!?== persons! as well as search sets designed to test
different aspects of face recognition methods.
Three /0T have been held *in ?===! ?==?! and most recently in ?==@+! where
researchers were invited to evaluate their algorithms andAor products against the competition.
,hile this process has been useful for researchers and possible buyers to see which methods
perform well and what their strengths and weaknesses are! it has also helped to define the
requirements for future work.
l6
/esearchers at 5olorado State University have collected most of the face recognition
methods that have performed well and distributed them free of charge. In the remainder of this
chapter! I first review the performance of the three main methods in this distribution! and then I
describe each of these three methods briefly.
4.4 I&%RO#.C%IO&
2iterature review is an activity of establishing a foundation for research by identifying other
works in a respective field. 2iterature review leads to problems! opportunities! and research
possibilities that have been overlooked by other researchers. 3ersonal e-perience! and
background! primary sources i.e. interviews! and secondary sources i.e. ,eb! books> can be
sources of conducting a literature review. In my case! I will look at literature review in the field
of access control through face recognition which can be established based on the applications or
implementations of the system in the global! regional! and local perspectives. The aim is to
identify opportunities which are to be leveraged and problems which can be solved in all the
three perspectives.
4.5 +$OBA$ $I%!RA%.R! R!/I!*
#lobally! facial recognition systems are deployed in large6scale citi)en identification applications!
surveillance applications! law enforcement applications such as identifying criminals! booking
stations! kiosks! computer access control! information system access control! and among many
other applications. 0arious approaches or algorithms to facial recognition have been in use
worldwide to implement such systems. erdous *?==B! p. :9 C ?:+ has e-plained various
methods in detail in his thesis and some of them include &igenfaces! 3rincipal 5omponent
"nalysis *35"+! 2inear 4iscriminant "nalysis *24"+! Support 0ector %achine *S0%+! 1iden
l7
%arkov %odels *1%%+! among many others. %ost of these methods use matrices and
probability techniques to e-tract and analy)e facial parts.
"ccording to the United States National Security Technology 5ouncil *NST5! ?==@+!
Subcommittee on biometrics! predominant approaches of face recognition problem include(
geometric *feature based+ and photometric *view based+. 4ue to increased demand on research
on the field! different algorithms were developed! in which three of them have been well studied
in face recognition literature( 35"! 24"! and &lastic Bunch #raph %atching *&B#%+.
2.3.1 Principal Component Analysis (PCA)
35" also called probabilistic eigenfaces are Dbased on the idea that you can appro-imately
rebuild a particular facial image using an average facial image along with information about
what differs from the averageE *B. %oghaddam! quoted from #uFnason! ?==@+. 3rincipal
component analysis *35"+ is a feature e-traction method! based on linear transformation
that has been used in data analysis pro$ects! e.g. ace recognition! audio recognition and
more. The main goals of 35" are two. The first is to identify patterns in data by ma-imi)ing
variance between components *patterns+ and the second is to reduce the dimensionality of the
data set while not losing much of the information. This reduction in dimensions removes
information that is not useful and precisely decomposes the face structure into orthogonal
*uncorrelated+ components which are known as eigenfaces. "n e-ample of eigenfaces can be
seen in figure ?.:(
l8
Figure 1 Standard !igenfaces3 Feature vectors are derived using eigenfaces
Source: %IT %edia 2aboratory Vision and Modeling Group: Photobook/Eigenfaces
Demo /etrieved online on ??
nd
ebruary ?=:? from
Ghttp(AAvismod.media.mit.eduAvismodAdemosAfacerecAbasic.htmlH
1owever! the 35" approach typically requires the full frontal face to be presented each
time the user wants to login> otherwise the detection and recognition results in poor performance.
2.3.2 Linear Desciminant Analysis
2inear discriminate analysis *24"+ also called isherfaces has been used in face recognition
systems! as well as other systems with comple- data! for data classification and
dimensionality reduction. The goal of 24" is to reduce the difference inside classes while
increasing the difference between them. In face recognition systems class represents person! but
l9
the class consists of many images per face. 'bviously it is a drawback of the method that it relies
on many images per person. Therefore Ihao et al. *quoted from #uFnason! ?==@+ combined the
2inear 4iscriminant "nalysis with the 35" algorithm to try to increase the recognition rate of
the faces using few images per person. The 35"J24" algorithm works in the following way.
irst it uses 35" *described in section ?.9.:+ to reduce the dimensionality of the feature
vectors and then 24" is run to produce a linear discriminant function that maps the input
into a classification space where the faces can be classified *,. Ihao> /. Beveridge! quoted
from #uFnason! ?==@+ by using some distance function. "n e-ample of fisherfaces can be seen
in figure ?.?(
Figure 2 !:ample of Si: Classes using $#A
Source( Kuwei 2u! *?==?+ oosting !inear Discriminant "nal#sis for $acial %ecognition.
2.3.3 Elastic Bunch Graph Matchin (EBGM)
The &lastic Bunch #raph %atching algorithm *&B#%+ is based on the Bochum A US5
*2auren) ,iskott! quoted from #uFnason! ?==@+ face recognition algorithm that was used in
the &/&T evaluation. That algorithm appeared to work well in the /0T test. The &B#%
algorithm performs
20
training on a set of facial images! where features are e-tracted from manually selected
points in the images. These points represent facial features such as left eye! right eye! nose!
etc. The e-tracted features are called model $ets and contain frequency information for the
selected points. Together the points and the model $ets are called landmarks. "ll the landmarks
for all the images in the training set are collected together into a graph called bunch graph. &very
node in the bunch graph represents facial features and contains landmarks for that particular
facial feature from all the images. The bunch graph can be used to locate landmarks in other
images *4. S. Bolme> K. /. Beveridge! quoted from #uFnason! ?==@+.
The &B#% method then creates a face graph for every facial image in a database. The
face graphs are similar to the bunch graph! e-cept that each face graph represents only
one facial image. These face graphs consist of many landmarks and for every landmark location
in the facial image a #abor $et is e-tracted! which consists of frequency information
about the pi-els surrounding the landmark locations. The #abor $ets are made by convolving
the landmark location with a collection of #abor wavelets and therefore they contain information
on the landmark and the nearest area surrounding it. The #abor $ets are based on L= comple-
wavelets where each wavelet has a real and imaginary component. These #abor wavelets are
similar to ourier analysis! but they cover only the point and its nearest neighborhood! so
frequency information far away from the point does not affect the wavelet. "n e-ample of
&B#% can be seen in figure ?.9(
2l
Figure 3 !lastic Bunc' +rap' Matc'ing
Source: 2auren) ,.!Kean. . Norbert MrNger! O 5hristoph %. *?==9+ $ace recognition b# elastic
bunch graph matching. /etrieved online on ??
nd
ebruary ?=:? from(
Ghttp(AAitb.biologie.hu6berlin.deAPwiskottA3ro$ectsA&#%ace/ecognition.htmlH
4.7 R!+IO&A$ $I%!RA%.R! R!/I!*
/egionally! acial recognition has been used especially in South "frica to provide access control
to buildings. T%"55 is an e-ample of a company based in South "frica offering biometric
solutions including facial recognition systems. "ccording to T%"55 case study! DIts business
philosophy revolves around preventing fraudulent activities such as unauthori)ed access to
restricted areas! preventing vandalism! industrial espionageE! among other threats. Its system
utili)es the 2ocal eature "nalysis *2"+ algorithm and 0ISI'NI5S technology for face
recognition.
1owever! better and more accurate algorithms have been developed in recent years such
as morphable 94 model! and neural6networks which outperforms the 2" algorithm especially
in uncontrolled scenarios. Therefore! taking advantage of the current technologies is significant
to improve face recognition applications.
22
4.8 $OCA$ $I%!RA%.R! R!/I!*
2ocally! facial recognition has not yet been identified as an industrial standard in the various fields as
e-plained in the global literature review. Such technology is mostly used at the individual level especially
for desktop login. But in Menya! facial recognition has been seeing potential as this article by I% &e"s
Africa recogni)es the future implementation of face recognition in "T% machines in order to do away
with 3IN and creditAdebit cards which can be stolen or duplicated. The author says that Dthe system uses a
camera that recogni)es the customer.s face using 94 biometrics *no life6si)e photographs can be used+!
sending details to a database and once verified! the customer is advised to enter the 3IN number and ask a
very personal question before using the "T% as usualE. 4r. %wangi ,aweru! the developer of the
system! also establishes that twins could pass the face test hence the 3IN or the question stage would be
the final test in this case.
S.MM!RY 9CO&C$.SIO&
'ut of the three approaches to face recognition described in section ?.9.:! ?.9.?! and ?.9.9! these
are among the commonly used in developing the current face detection and recognition systems
and most widely documented around the world. I am proposing the use of 2aplacianfaces
approach developed by Xiaofei He and his colleagues in 2005 while at State Mey 2ab of
5"4O5#! Ihe$iang University! in 1ang)hou 5hina. This algorithm makes use of 2ocality
3reserving 3ro$ections to improve the accuracy of detection and recognition as compared to the
other three approaches discussed earlier in this chapter.
Theoretically! based on tests carried out in Qale! 3I& and %S/" face databases! this
approach can increase the recognition performance especially in variations of face angles! facial
e-pressions! pose! lightning conditions! and among other variations in uncontrolled conditions by
appro-imately ?RS through comparing error rates and dimensions of face subspace in various
23
databases. 1owever the actual implementation of this approach has not been reali)ed in day to
day applications. 1ence! the aim of this research is to see how this approach can be e-ploited to
solve the problem of automated face recognition while improving the accuracy level especially
in uncontrolled conditions.
C0A1%!R 53 M!%0O#O$O+Y
This chapter focuses on identifying the various beneficiaries of the study! the various methods
used for sampling! data collection! data analysis! system analysis and design! and implementation
language. It also provides a simple test plan for the system.
5.2 I&%RO#.C%IO&
Today.s world is moving much towards information technology which requires more and more
systems to handle or rather simplify our day to day operations as well as provide secure means of
data storage. %ethodology aims at identifying the beneficiary of the research including the
sample group> identifying the various methods to be used in data collection> analy)e the system
24
functionality and interaction of various components through system analysis and design
methods> pro$ect scheduling using #antt chart> and evaluation of operation of various modules
and the entire system through various stages of tests.
5.4 SAM1$! ;%AR+!% +RO.1<
The facial recognition system will be designed to run as a prototype in an information system for
administration of user accounts. The system will run through a parallel strategy as I analy)e its
efficiency and accuracy. 3arallel strategy involves making use of the facial recognition system as
well as the legacy means of authentication that is usernames and passwords to gain access to an
information system. This will prevent the user to be locked off from his account if the system
fails to recogni)e the user.
"bout ?= students. sets of face samples will be used in testing the system accuracy and
efficiency. The system is required to log the students in their respective accounts as well as
performing other functionalities as mentioned in the system ob$ectives in chapter :. If the
evaluation shows positive results! the system is intended to be open source where other
developers can improve on the system as well as integrate the system to other applications
requiring sophisticated and improved means of authentication.
5.5 SAM1$I&+ M!%0O#
/andom sampling method will be used to select the ?= students who are willing to participate in
providing feedback concerning the prospects of the study. The same students will be used during
system testing. This sampling technique is easier to carry out since the sub$ects are not limited by
any form of demographic characteristics such as gender! age! and among others which is the case
25
with stratified sampling. "ny student or individual can participate since the research aims at
influencing how people gain access to various systems they interact with every single day.
5.7 #A%A CO$$!C%IO& M!%0O#;S<
This shows the types of data collection procedures that will be used to gather data regarding the
system. I will use the following data collection methods(
I. 4ocument e-ploration
II. ield intervention
III. Tuestionnaires
In each of them there are different research instruments and data collection procedures
that are going to be followed in carrying out the research.
5.7.2 #ocument e:ploration
4ocument e-ploration is the process of going through a number of literary works that have been
conducted by other authors to review their thoughts! ideas! and opinions towards a particular
sub$ect. It includes the following instruments(
i. %eading from the librar#
In library readings! I will be able to come up with ideas of developing the system and methods of
classifying data including an in depth understanding of various mathematical algorithms and
sub$ects such as probabilities and matrices that will be useful during system implementation. It
will also help me in coming up with new ideas on how to structure the system.
ii. %esearch from the &nternet
26
This is where I will go online and gather as much information about related systems and their
implementation. This will enable me to open my mind and embrace new ideas.
iii. Exploring digital libraries
This will ultimately involve the use of the Internet to access the online libraries that has
collection of various articles related to the development of the automated facial recognition
system.
5.7.4 Field Intervention
ield intervention is where an individual goes to e-amine a sub$ect in its naturally occurring
environment rather than reading literary works or performing e-periments. This is done basically
through conducting interviews. Sample questions can be found in appendi- II.
5.7.5 =uestionnaires
This is an alternative to field intervention where instead of a researcher going to the field to
conduct interviews! the interview questions are printed out on paper. &very sub$ect receives a
copy of the questions and provides feedback after a particular period of time. It is a convenient
way of data collection! but sometimes sub$ects might be bias towards the questionnaires hence
delivering une-pected low quality feedback. In my case! the same questions used for field
intervention will be used in questionnaires.
5.8 SYS%!M A&A$YSIS A&# #!SI+& M!%0O#;S<
S425 also called the System 4evelopment 2ife 5ycle is the system development methodology
used to plan! analy)e! design! implement! and support the information system. The method
usually includes five *R+ ma$or steps as outlined below(
27
i. Systems planning phase
ii. Systems analysis phase
iii. Systems design phase
iv. Systems implementation phase
v. Systems operation! support! and security phase
The end product of the system development method is the system that can be delivered
and become useful to the intended users who were discussed in the beneficiary and the target
group sections.
System analysis and design methods used include the 4ata low 4iagrams *44s+! Use
5ases! Sequence 4iagrams! and &ntity /elationship 4iagrams. This has also included database
schema structure for the face database.
5.> CO#I&+ ;$A&+.A+! %O B! .S!#<
The programming language used in implementing the entire system is Kava. Both the facial
recognition module along with all its features as well as the information system has been built on
Kava. The Netbeans I4& < has been used in designing the various sub6systems and modules and
Kava implementation for the coding part. %icrosoft ,indows < @Lbit has been used as the
implementation platform. 1owever! the system has also been implemented in any platform with
Kava /untime &nvironment @ support.
5.? %!S%I&+ 1$A& FOR %0! SYS%!M
The system will be tested through three ma$or steps which include unitAmodule testing!
performance testing! and to finish with system testing.
28
.nit %esting( It involves testing of individual software components or modules. In my case!
testing has been carried out in the enrollment! detection! and matching modules. "lso! testing has been
done on user accounts management module.
1erformance %esting( Testing the efficiency and accuracy of detection and recognition.
System %esting( Testing the overall system functionality making sure all modules work together.
C0A1%!R 73 SYS%!M A&A$YSIS A&# #!SI+&
This chapter focuses on various systems analysis and design techniques that will help us to
visuali)e how the system will function after implementation on the basis of its internal
operations! its interaction with users! and the output e-pected from the system. Some of the
techniques used include use cases! activity diagrams! data flow diagrams! class diagrams! and
entity relationship diagram. But before this! we should first hand understand the functional and
non functional requirements for the system in order to form the basis for system design.
7.2 F.&C%IO&A$ R!=.IR!M!&%S
29
In order for the system to accept the right data! process the data! store the data! and provide the
right output when needed! functional requirements for the system have to be satisfied.
"utomated face recognition system has the following functional requirements(
I. The administrator should be able to create! view and delete user accounts with relevant
user records.
II. &ach record represents information about a person and contains an image of hisAher face.
III. /ecords may consists of(
i. irst name.
ii. 2ast name.
iii. 3hone.
iv. "ddress *city6street address+.
v. ace image*s+.
vi. I46number.
I0. The system is able to take face image as an input from a webcam and search for a
matching face in the database! and then show the results.
0. The results can be viewed by showing the faces matches the input! and logs the user into
the information system.s relevant user account.
0I. The user can be able to view hisAher full record.s information once heAshe gains access to
the information system.
30
0II. The system can also do noisy filter on images.
7.4 &O& F.&C%IO&A$ R!=.IR!M!&%S
In order for the functional requirements to be achieved! the non functional requirements are
important as they provide the foundation for the actual operation of the system. The non
functional requirements for the system include(
I. /ecords are maintained in a database.
II. &very record is allocated a unique identifier *id6number+.
III. The administrator is able to retrieve user data by entering id or name.
I0. The images are pre6processed by ignoring the irrelevant parts! means locali)ing
and normali)ing the face.
0. The face can be locali)ed by detecting inner and outer boundaries! background must be
ignored.
0I. To improve the performance! the database include a transformed image *template+ by
233 algorithm for each record! i.e. the template should be ready when it.s needed.
0II. The database is developed using %icrosoft ST2 ServerA%yST2.
7.5 .se Case #iagrams
These are diagrams that visuali)e the relationship or interaction between users of the system and
the system itself. " use case is made up of actors which are users! associations which are the
relationships! and cases which are the activities or operations performed by the user or invoked
3l
by a module in the system. In my case! the automated face recognition system has only two
actors who are the user *who has low access level+ and the administrator *who has high access
level+. The complete picture can be shown in the following figures(
32
Figure 4 .ser addition and removal scenario
The first scenario is called the User addition and removal scenario and is depicted in ig. L. In
this scenario there are only two types of actors or users that will interact and use the system.
These are mainly an "dministrator and a User. The "dministrator is the system user that will add
and remove other Users that must be recogni)ed by the system through a graphical user interface.
The User is the system user that is to be added for recognition by the "dministrator and interacts
with the system through a webcam.
33
Figure 5 .ser face recognition scenario
The second scenario is called the User ace /ecognition Scenario and is depicted in ig.
R. In this scenario! the system is used to recogni)e a User that was added! to be recogni)ed by the
system by using a webcam and a graphical user interface that contains a live video window. The
window is used by the User to help position the Users face for the images that are to be captured
by the system.
7.7 Activity #iagrams
"n activity diagram is a model that shows step by step activities of the user in different
scenarios. It helps in coming up with efficient graphical user interfaces in which users interact
with the system. In my case! activity diagrams range from both the administrator and the user as
can be seen in the figures below(
34
Figure 6 Insert ne" user Activity by t'e administrator
35
Figure 7 Re@uest Matc'ing9$ogin Activity by t'e user or t'e administrator
Figure 8 Remove user9user account Activity by t'e administrator
The first activity is called the insert new user activity and is depicted in figure @. In this
case! the administrator can enroll both the personal details and the face samples of the user. The
administrator will open an account form where a record is entered. " camera stream will be used
to capture the face samples. If the information concerning a particular user already e-ist! the
system will display an error message.
The second activity is called the request matching activity and is depicted in figure <. In
this case! the user or administrator gains access to the system through facial recognition. The
user will access the webcam to enable a video stream. The face recognition algorithm is used to
match the face in the stream with the face samples. If there is no match! the user can not gain
access to the information system.
The third activity is called the remove user activity and is depicted in figure B. In this
case! the administrator is capable of deleting user accounts based on their I4s. " user I4 is
provided to the system and the delete request is initiali)ed. If the user e-ist! the system will
remove his record! but if not! the system will display an error message that the user account does
not e-ist.
7.8 #ata Flo" #iagrams ;#F#s<
These are diagrams that show the flow of data among the system modules! sub systems! users!
and e-ternal systems. They are made up of entities which hold the e-ternal parts of the system
such as the users and e-ternal systems! processes which are the functional parts of the system!
data flows to show movements of data! and data stores for holding data in form of records. In my
36
case! 44s for the automated facial recognition system include 44 conte-t *level =+ and 44
level : diagrams as shown in the following figures(
37
Figure 9 System #F# Conte:t #iagram
38
Figure 10 System #F# $evel 2 #iagram
rom figure ; which shows the 44 level = 5onte-t 4iagram! it is made up of three
entities i.e. User! "dministrator! and the IS. The IS is not part of the /S which is the main
process in this case. Both the user and the administrator request access to the IS through the /S
where there can be an acceptance or re$ection from the system. The /S also provides auto6
authentication functionality to improve security of the IS.
igure := shows the 44 level : which is made up multiple processes i.e. take snapshot!
face detection! feature e-traction! face matching! add! view! and delete accounts. ace images
and user accounts form the data stores for the system while userAadministrator and the IS form
the entities. The user access a webcam where a video stream is loaded. ace detection and
feature e-traction algorithms are used to differentiate a face from other ob$ects from the video
stream. " face image is produced in the process where a face recognition algorithm is used to
calculate a unique covariance matri- of the face sample. %ultiple face samples of the same user
will create an average covariance matri- which improves recognition. The user is authenticated if
there is a match! hence if he has administrative privileges! adding! viewing! and deleting
activities can be done to the IS.
7.> Class #iagrams
These are U%2 diagrams that represent the relationships among classes making up a system.
5lasses contain functionalities of various modules or sub6systems. In my case! there are various
classes making up the system functionality including libraries imported from e-ternal sources!
user interfaces! database connection settings! matri- engine for vector calculations! main
methods! and 2aplacian attributes and methods.
39
40
Figure 11 Classes #escription class diagram
4l
Figure 12 Relations'ip Bet"een Classes class diagram
7.? !ntity Relations'ip #iagram
It is a modeling technique that shows the relationship between various entities within a database.
It is used to present a conceptual model of data that is how data will be stored and accessed in the
database. It is made up of entities! relations! and particular types of data stored. In my case! the
&/4 is mostly made up of one table within a database holding user details along with hisAher
image. "n entity relationship diagram for the automated facial recognition system is shown in
the figure below(
Based on the &/4! N means many which show a user or administrator can have multiple
face samples in the database where the face samples will have image numbers attribute.
C0A1%!R 83 *OR) 1$A& A&# B.#+!%
42
.S!R9
A#MI&IS%RA%OR
IMA+!
lD
fname
lname
age
address
lmage #
phone
Ha
s a
: N
Figure 13 System !ntity Relations'ip #iagram ;!R#<
This chapter describes the scheduling of the pro$ect in which various pro$ect tasks! under the
main ob$ectives! are placed in various durations. It also gives out the resources or tools used in
carrying out the various ob$ectives and tasks along with their quantities and costs.
8.2 *or- 1lan
The pro$ect is e-pected to take appro-imately ?? weeks in total. There are a total of :9 tasks that
upon completion will lead to the accomplishment of the pro$ect goal. The tasks have been
grouped under each of the pro$ect ob$ectives. This is as shown in the following tables with
ob$ectives : through R.
'()() Pro!ect Goal
The purpose of this research study is to develop a facial recognition system that will keep track
of the authenticity of the user of a particular computer system or particular information system
and to improve the accuracy of recognition especially in uncontrolled conditions involving
variations in lightning! facial e-pressions! facial angles! and pose by implementing a
2aplacianface approach with 2ocality 3reserving 3ro$ections *233+ to facial recognition.
'()(* Pro!ect "#!ecti$es an% &trateies 'or Achie$in (hem
Table 1 obAective 23 to develop user re@uirements91reliminary investigation
ObAective 2 #evelop t'e 1reliminary Investigation doc #uration 8 "ee-s
%AS) S%AR% FI&IS0 Action to %a-e Met'ods or tools9
Resources re@uired
!:pected deliverables
4ata 5ollection ?RA:A:? ?@A?A:? 4eveloping
Tuestionnaires and
Interview Tuestions.
%s ,ord.
"rticles.
Be familiar with the
system functionalities
43
3erforming document
e-ploration.
Books.
&6books
and user requirements
4ata "nalysis ?<A?A:? RA9A:? Transferred the data to
S3SS and analy)ed
S3SS ! %oS5o, #aining familiarity with
S3SS
Table 2 obAective 43 to design t'e system and database re@uirements
ObAective 4 %o design t'e system and database
re@uirements
#uration 4 "ee-s
%AS) S%AR% FI&IS0 Action to %a-e Met'od or tools9
Resources re@uired
!:pected deliverables
U%2 modeling @A9A:? :=A9A:? 4esigning use cases!
activity! and class
diagrams
Unified %odeling
2anguage in U%2&T
U%2 models
Business
process
modeling
:?A9A:? :RA9A:? 4esigning data flow
diagrams
4ata low 4iagrams
*44s+ in 'pen
%odelSphere
44s models
44
4ata modeling :@A9A:? ?=A9A:? 4esigning the user and
administrator table! and
established the
relationships.
&ntity /elationship
4iagrams *&/4s+ in
'pen %odelSphere
&/4s models
Table 3 obAective 53 to implement system functionalities
ObAective 5 %o implement system functionalities #uration 24 "ee-s
%AS) S%AR% FI&IS0 Action to %a-e Met'od or
tools9Resources
re@uired
!:pected deliverables
4atabase
creation
?:A9A:? ?=ALA:? 5reating user and
administrator table! and
established the
relationships.
%icrosoft ST2 server
?==BA%yST2
4atabase schemas
User Interface
modeling and
designing
?=ALA:? BA@A:? 5reating login!
enrollment! and user
accounts management
interfaces.
NetBeans I4& < and
5"S& tools
System #raphical User
Interfaces
System coding ?=ALA:? 9A<A:? 5oding various system
modules i.e. enrollment!
detection! and matching>
user accounts
management> login> etc.
K%! %atri- 2ibrary!
"orge 2ibrary! and
NetBeans I4& <
System modules and
subsystems
Table 4 obAective 73 to carry out testing and maintenance
45
ObAective 7 %o carry out system testing #uration 4 "ee-s
%AS) S%AR% FI&IS0 Action to %a-e Met'od or
tools9Resources
re@uired
!:pected deliverables
%oduleAUnit
Testing
LA<A:? ;A<A:? Testing the enrollment!
detection! and matching
modules. Testing User
accounts management
module.
Trial runs "ll modules work
accordingly
3erformance
Testing
:=A<A:? :?A<A:? Testing the efficiency
and accuracy of
detection and
recognition.
Trial runs #ood performance
System Testing :9A<A:? ?=A<A:? Testing the overall
system functionality
making sure all modules
work together.
Trial runs The entire system work
accordingly
Table 5 obAective 83 to finaliBe documentation and presentation
ObAective 8 %o finaliBe documentation
and presentation
#uration 2 "ee-
%AS) S%AR% FI&IS0 Action to %a-e Met'od or
tools9Resources
re@uired
!:pected deliverables
4ocumentation
review
?:A<A:? ?@A<A:? inali)ing the
documentation and
%S ,ord " complete and final
46
evaluating the different
chapters
system documentation
3resentation
development
?:A<A:? ?@A<A:? inali)ing the
presentation
development
%S 3ower3oint " complete and final
3ower3oint presentation
for the system
8.4 Budget
The cost of the proposed system is shown in table @ below. It is showing in detail the item!
quantity! and its cost.
Table 6 t'e cost estimation of various resources to be used in t'e researc' proAect
I%!M #!SCRI1%IO& =.A&%I%Y C)S0 %O%A$
AMO.&%;)S0<
2aptop 5omputer
with ,ebcam
System platform for
modeling!
implementation! and
testing.
: <=!=== <=!===
"NU Internet 5ost of internet usage
while doing preliminary
investigation and system
development.
? trimesters L=== B===
Transport /ongai 6 Town 9 :== 9==
47
3rinting ee 4ocumentation 9 ?== @==
"irtime Safaricom broadband ? R== :!===
Binding ee or the documentation 9 R= :R=
%otal DEE8E
CO&C$.SIO&
ace recognition technology brings in new opportunities into the field of biometrics as machines
are given the capability of identifying human faces like normal human beings. This has
contributed to the development of human like machines also known as humanoid robots. The
ma$or breakthrough of the technology came when the technology was applied in authenticating
users. This has led to increasing the security and privacy of users. data as well as improvement
of buildings security. It has also become an alternative to the traditional means of access control.
In Menya! we need to embrace this technology in order to have a technology driven society that
take an advantage of sophisticated technologies around us.
48
R!F!R!&C!S
IT News "frica. *?=:=+ +en#a could implement face recognition technolog# for ",Ms. /etrieved
'nline on ?B
th
Kanuary ?=:? from( Ghttp(AAwww.itnewsafrica.comA?=:=A=RAkenya6could6
implement6face6recognition6technology6for6atmsAH
erdous I. S. *?==B+ !iterature Sur-e# on "utomatic $ace %ecognition S#stem and Eigenface
based implementation. /etrieved online on 9=
th
Kanuary ?=:? from(
Ghttp(AAdspace.bracu.ac.bdAbitstreamAhandleA:=9@:A:@;A2itelatureS?=survayS?=of
S?=automaticS?=faceS?=recognitionS?=system.pdfUsequenceV:H
#uFnason 1afWXr *?==@+ Median %ank in $ace %ecognition. /etrieved online on :;
th
ebruary
?=:? from( Ghttp(AAwww.ru.isAlisalibAgetfile.asp-UitemidV;RR:H
Kuwei 2u! *?==?+ oosting !inear Discriminant "nal#sis for $acial %ecognition
2auren) ,.!Kean. . Norbert MrNger! O 5hristoph %. *?==9+ $ace recognition b# elastic bunch
graph matching. /etrieved online on ??
nd
ebruary ?=:? from(
Ghttp(AAitb.biologie.hu6berlin.deAPwiskottA3ro$ectsA&#%ace/ecognition.htmlH
49
%IT %edia 2aboratory Vision and Modeling Group: Photobook/Eigenfaces Demo /etrieved
online on ??
nd
ebruary ?=:? from
Ghttp(AAvismod.media.mit.eduAvismodAdemosAfacerecAbasic.htmlH
United States National Security Technology 5ouncil *NST5+! Subcommittee on biometrics
*?==@+ $ace %ecognition: " case stud# /etrieved online on :;
th
ebruary ?=:? from(
Ghttp(AAwww.biometrics.govA4ocumentsAfacerec.pdfH
BIB$IO+RA10Y
ace /ecognition 'rgani)ation *?=:?+. /etrieved online on 9:
st
Kanuary from( Ghttp(AAwww.face6
rec.orgH
1e! Yiaofe> Qan! Shuicheng> Niyogi! 3artha and Ihang! 1ongKiang *?==R+. $ace %ecognition
.sing !aplacianfaces /etrieved online on 9=
th
Kanuary ?=:? from(
Ghttp(AApeople.cs.uchicago.eduAPniyogiApaperspsA2aplacianface.pdfH
1offer ". K.! #eorge . K.! O 0alaaich S. K. modern s#stem anal#sis and design *@
th
&d.+. New
Qork( 3earson
Mendall &. M.! O Mendall &. K. s#stem anal#sis and design *B
th
&d.+. New Qork( 3earson
3! Belhumeur and K! 1espanha *:;;<+. &igenfaces vs. isherfaces( %ecognition .sing /lass
Specific !inear Pro0ection: &EEE ,ransactions on Pattern "nal#sis and Machine
&ntelligence
50
Q.! 5hang> 1u! 5. and Turk! %. *?==9+. Manifold of $acial Expression. I&&& Int8l ,orkshop
"nalysis and %odeling of aces and #estures! ?==9
Y. 4. Qang! 3. Mort! and /. 4osselmann! D"utomatically log off upon disappearance of facial
image!E 5ontract /eport 5/ ?==R6=R:! 4/45! 'ttawa! 5anada! %arch ?==R.
A11!&#IF I3 +A&%% C0AR%
5l
A11!&#IF II3 SAM1$! =.!S%IO&S
:. 1ow do you find the current username or email and password system of gaining access to
various websites! systems! or your own desktop or laptop computerU
?. 4o you need alternative means of gaining access to these systemsU Q&S or N'U
9. 1ave you ever heard of face recognitionU Q&S or N'U
L. If Q&S! 3lease describe your understanding on face recognitionU
R. 1ave you ever used face recognition in any wayU Q&S or N'U
@. If Q&S! what systems have you applied face recognitionU
52
A11!&#IF III3 SAM1$! CO#!S
Face %rac-ing Met'od3
private void trackFace(final Bufferedlmage img)
/¯ Create a separate thread for the time-consuming detection and recognition tasks:
find a face in the current image, store its coordinates in faceRect, then
recognize the face, and draw the person's name in the panel
¯/
{
final Bufferedlmage graylm = lmageUtils.toScaledGray(img, l.0/lM_SCALE); // reduce size
numTasks.getAndlncrement(); // increment no. of tasks before entering queue
executor.execute(new Runnable() {
public void run()
{
detectStartTime = System.currentTimeMillis();
CvRect rect = findFace(graylm);
if (rect != null) {
setRectangle(rect);
if (recognizeFace) {
recogFace(img);
53
recognizeFace = false;
//faceName="";
}
}
long detectDuration = System.currentTimeMillis() - detectStartTime;
System.out.println(" detection/recognition duration: " + detectDuration + "ms");
System.out.println(" Recognized Face lD: " + faceName);
numTasks.getAndDecrement(); // decrement no. of tasks since finished
}
});
} // end of trackFace()
Auto,Recognition Met'od3
public void autoRecognition(){
/¯ handles continous recognition of a person while operating the system.
¯ Performs auto loggoff if the person is not authenticated
¯/
int sessionlD = l;
if (faceName != null){
DBConnection con = new DBConnection(); // creating a database instance
//comparing user input to the database
String credentials="SELECT ¯ FROM session WHERE sessionid='"+sessionlD+"' AND
accountid = '"+faceName+"'";
con.Retrieve(credentials);
try {
String FirstName = null;
String User_Level = null;
String Account_lD = null;
//String Staffpass = null;
if (con.rs.next()) {
//User_Level = con.rs.getString(8);
//Account_lD = con.rs.getString(l);
//FirstName = con.rs.getString(3);
}
else {/¯
int i=0;
while(i<l0){
Toolkit.getDefaultToolkit().beep();
i++;
}
¯/
RepeatBeeps();
JOptionPane.showMessageDialog(this,"No match found for the detected face! The
system will loggoff!","Error",JOptionPane.ERROR_MESSAGE);
if (faceView != null){
//initializing empty variables
String lD = "";
String Level = "";
//updates the session values to the database
54
String sql = "UPDATE session SET accountid='"+lD+"', accountlevel='"+Level+"'";
con.UpDate(sql); // updates the database
faceView.dispose();
new FaceAccessLoginControl().setVisible(true);
}
}
} catch (SQLException ex) {
Logger.getLogger(FaceRecogPanel.class.getName()).log(Level.SEVERE, null, ex);
}

finally {
}
}
else{/¯
int i=0;
while(i<l0){
Toolkit.getDefaultToolkit().beep();
i++;
}
¯
¯/
RepeatBeeps();
JOptionPane.showMessageDialog(this,"No face to recognize! The system will
loggoff!","Error",JOptionPane.ERROR_MESSAGE);
if (faceView != null){
DBConnection con = new DBConnection(); // creating a database instance
//initializing empty variables
String lD = "";
String Level = "";
//updates the session values to the database
String sql = "UPDATE session SET accountid='"+lD+"', accountlevel='"+Level+"'";
con.UpDate(sql); // updates the database
faceView.dispose();
new FaceAccessLoginControl().setVisible(true);

}
}

} // end of autoRecognition()
55
A11!&#IF I/3 SAM1$! SCR!!&S0O%S
%ain Screen
56
This shows the main window of the application. It gives a user access to various system tasks
like face enrollment! user profile! configuration! and among others.
ace &nrollment
57
This is the main face enrollment window where user.s face samples *templates+ are captured.
The system uses this information to determine the identity of a particular user during facial
recognition.
58
Two6actor "uthentication(
This is a two6factor authentication window that enables a user to login with the password and I4
as well as facial recognition. It provides a more secured access control mechanism for user
accounts.
59

Sponsor Documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close