Table Of Contents

Topic Page No.

1. Introduction, Objective & Scope 2-3

2. Review of Literature 4

3. Materials & Methods (Technical Details) 5-10

3.1 Project Category 5

3.2 Techniques To Be Used 5-7

3.3 Hardware and Software Resource Requirements and their Specifications 7-10

4. Proposed Methodology 11-13

4.1 Proposed Algorithm/ Software Code 11

4.2 System Architecture & Flow Chart 11-12

4.3 Software Implementation 12-13

5. Testing Technologies and Security Mechanisms 14-15

6. Future scope and Limitations 16

7. Conclusion 17

8. Bibliography 18

1. References 18

1
 Introduction, Objective & Scope

In today’s world security is very much important for the well being of human and to

overcome from any crime. When we talk about security the first thing that comes in our mind

is the video surveillance. When looking at video surveillance systems, you will most likely

come across the term CCTV. CCTV stands for Closed-Circuit TeleVision and refers mainly to
the idea that the video footage is sent over a single channel with a dedicated set of wires
(circuit) feeding into a monitor or recording device – creating what is essentially a closed
circuit. Rather than being broadcast publicly for all to see, any cameras, monitors or recorders
used are communicating via a single wired or wireless connection, with no other outside
connections or observers.
In China, government uses “Face Recognition CCTV Camera System” to detect criminals.
They use CCTV surveillance system that are fitted in the street and had a face recognition
system that detect the face of the person that face data has been stored in the database of the
system. The hardware part consists of a CCTV camera, while the software part consists of
face-detection and face-recognition algorithms software. The camera detects the face for
which it has been designed at a particular time it alerts the nearby police station that the
particular person has been found at that locality. Their “Face Recognition CCTV Camera
System” are used to detect the face of the person which is already stored in the database and
alert the nearby police station but these CCTV camera costs a lot more expensive which is
not affordable by everyone so we are going to make some changes and build it up at a low
cost comparatively to those of China CCTV camera system.
We are going to build up the “Face Recognition CCTV Camera” which has been used to
detect an unfamiliar face which has been not stored in the database of the security system.
Here we are going to stored the faces of the people who are allowed to come where this
security system going to be installed.
In the range of camera zone, when a person will enter a series of snapshots will be taken by
the camera and will send it to the software to be analyse and compare with an existing
database of trusted people and if the person will not be recognised then the system will send
an SMS to the registered user whose mobile number has been stored inside the system that

2
“An intrusion has been detected in the zone where the camera has been installed”. And it will
also start the burglar alarm that has been the integral part of this security system.

The main aim of this project is to detect intrusion and to increase the security in isolated area.
The main idea of this project is to detect the intruders where there is limited number of
people are allowed to enter i.e bank lockers, scientist research lab etc. It is very much cost
effective rather than the normal Face Recognition CCTV Camera available in the market as
well as it has a GSM module to alert the owner which is not available in normal Face
Recognition CCTV Camera.

3
 Review of Literature

• Face Recognition Security System (CISSE 2013)

This paper gives high importance on security using face recognition system. It gives
us many features on software along with hardware to use facial recognition system for
security.

• Security System Based On Facial Recognition Using Arduino (IJAREEIE 2017)

This paper deals with the advance security system in automotive using face
recognition system with arduino.

• Face Search In CCTV Surveillance (Springer 2019)

This paper present a series of experiments on visual search in a highly complex

environment, security closed-circuit television (CCTV). Using real surveillance footage from
a large city transport hub, in which viewers are able to search for target individuals from the
video of CCTV.

4
 Materials & Methods

3.1 Project Category

Our proposed work is a software embedded module consists of various units that
make up the system: the power supply unit, Raspberry Pi 4, buzzer unit. Raspberry Pi 4 is
used to keep looking for the output from the infrared CCTV Camera and it will match the
faces that falls in the range of CCTV Camera. There is a built in GSM module which has
been developed using python API will used to send text message as soon as the unknown face
detects by the CCTV. After sending the text message it will start the buzzer alarm that will be
attached to the micro-controller. The Arduino Uno sketch which is the environment for
programming is used to write the code, compile, generate hex file and load it to the micro-
controller.

3.2 Techniques To Be Used

Face Recognition :

There has been a rapid development of the reliable face recognition algorithms in the
last decade. The traditional face recognition algorithms can be categorised into two
categories: holistic features and local feature approaches. The holistic group can be
additionally divided into linear and nonlinear projection methods. Many applications have
shown good results of the linear projection appearance-based methods such as principal
component analysis (PCA), independent component analysis (ICA), linear discriminate
analysis (LDA) , 2DPCA and linear regression classifier (LRC). However, due to large
variations in illumination conditions, facial expression and other factors, these methods may
fail to adequately represent the faces. The main reason is that the face patterns lie on a
complex nonlinear and non-convex manifold in the high-dimensional space.

In order to deal with such cases, nonlinear extensions have been proposed like kernel PCA
(KPCA), kernel LDA (KLDA) or locally linear embedding (LLE). The most nonlinear
methods using the kernel techniques, where the general idea consists of mapping the input
face images into a higher-dimensional space in which the manifold of the faces is linear and

5
simplified. So the traditional linear methods can be applied. Although PCA, LDA and LRC
are considered as linear subspace learning algorithms, it is notable that PCA and LDA
methods focus on the global structure of the Euclidean space, whereas LRC approach focuses
on local structure of the manifold.

These methods project face onto a linear subspace spanned by the eigenface images. The
distance from face space is the orthogonal distance to the plane, whereas the distance in face
space is the distance along the plane from the mean image. These both distances can be
turned into Mahalanobis distances and given probabilistic interpretations.

Following these, there have been developed: KPCA, kernel ICA and generalised linear
discriminant analysis.

Despite strong theoretical foundation of kernel-based methods, the practical application of

these methods in face recognition problems, however, does not produce a significant
improvement compared with linear methods. Another family of nonlinear projection methods
has been introduced. They inherited the simplicity from the linear methods and the ability to
deal with complex data from the nonlinear ones. Among these methods, it is worth to
underline: LLE and locality preserving projection (LPP). They produce a projection scheme
for training data only, but their capability to project new data items is questionable.

In the second category, local appearance features have certain advantages over holistic
features. These methods are more stable to local changes such as expression, occlusion and
misalignment. The common representative method names local binary patterns (LBPs). The
neighbouring changes around the central pixel in a simple but effective way are described by
LBPH. It is invariant monotonic intensity transformation and supports small illumination
variations. Many LBP variants are proposed to improve the original LBP such as histogram
of Gabor phase patterns and local binary pattern histogram sequence. Generally, the LBP is
utilised to model the neighbouring relationship jointly in spatial, frequency and orientation
domains.

6
It allows to explore efficiently discriminant and robust information in the pattern. Further
development of the mentioned subspace approaches represents discriminant common vectors
(DCVs) approach. The DCV method collects the similarities among the elements in the same
class and drops their dissimilarities. Thus, each class can be represented by a common vector
computed from the within scatter matrix.

In case of testing an unknown face, the corresponding feature vector is computed and
associated to the class with the nearest common vector. Sometimes, kernel discriminative
common vectors or improved discriminative common vectors and support vector machine
(SVM) are introduced for the face recognition task.

Along with the face recognition it is not forget that we need such an algorithm which can
detect human face from the input of camera, so we will using Viola–Jones object detection
framework. It is mainly developed to detect the human faces.

The Viola-Jones algorithm first detects the face on the grayscale image and then finds the
location on the colored image. Viola-Jones outlines a box (as you can see on the right) and
searches for a face within the box. It is essentially searching for these haar-like features.

3.3 Hardware and Software Resource Requirements and their Specifications

3.3.1 Hardware Requirements:

I. Raspberry Pi 4 (Model B)

The Raspberry Pi 4 Model B is the latest version of the low cost Raspberry Pi computer.
The Pi isn't like your typical device; in its cheapest form it doesn't have a case, and is simply
a credit-card sized electronic board of the type you might find inside a PC or laptop, but
much smaller.The Raspberry Pi 4 can do a surprising amount. Amateur tech enthusiasts use Pi
boards as media centers, file servers, retro games consoles, routers, and network-level ad-
blockers, for starters. However that is just a taste of what's possible. We will use it for image
processing.

7
II. Buzzer Unit

The buzzer unit which is used as an alarm which indicates when unknown face is detected.
The buzzer used belongs to the PS series. The PS series are high-performance buzzers that
employ Uni-morph piezoelectric elements and are designed for easy incorporation into
various circuits. They have very low power consumption in comparison to electromagnetic
units. It has a voltage requirement of 2V and is connected to pin 10 of the micro-controller.

III. Infrared CCTV Camera Unit

Infrared security cameras have the capability to capture video in low light and no light (0
Lux) areas. Infrared cameras (IR) are sometimes referred to as IR cameras or “Night
Vision Security Cameras”. This unit is used to record the faces which will come in the range
of CCTV camera and it will send the capture data (faces) to arduino micro-controller so that
the faces are match with the stored faces from the database.

3.3.2 Software Requirements:

I. Python 3

Python is an interpreted, high-level, general-purpose programming language. Created

by Guido van Rossum and first released in 1991, Python's design philosophy
emphasizes code readability with its notable use of significant whitespace. Its language
constructs and object-oriented approach aim to help programmers write clear, logical code for
small and large-scale projects. Python is dynamically typed and garbage-collected. It supports
multiple programming paradigms, including structured (particularly, procedural), object-
oriented, and functional programming. Python is often described as a "batteries included"
language due to its comprehensive standard library.

II. Anaconda Navigator

Anaconda is a free and open-source distribution of the Python and R programming

languages for scientific computing (data science, machine learning applications, large-scale

8
data processing, predictive analytics, etc.), that aims to simplify package management and
deployment. The distribution includes data-science packages suitable for Windows, Linux,
and macOS. It is developed and maintained by Anaconda, Inc., which was founded by Peter
Wang and Travis Oliphant in 2012. As an Anaconda, Inc. product, it is also known
as Anaconda Distribution or Anaconda Individual Edition, while other products from the
company are Anaconda Team Edition and Anaconda Enterprise Edition, which are both not
free.

III. VS Code

Vi s u a l S t u d i o C o d e i s a s o u r c e - c o d e e d i t o r d e v e l o p e d
by Microsoft for Windows, Linux and Mac OSX. It includes embedded Git and support
for debugging, syntax highlighting, intelligent code completion, snippets, and code
refactoring. It is highly customizable, allowing users to change the theme, keyboard
shortcuts, preferences, and install extensions that add additional functionality. The source
code is free and open-source, released under the permissive MIT License. The compiled
binaries are freeware for any use.

IV. NumPy With OpenCV 2 Library

NumPy is a package in Python used for Scientific Computing. NumPy package is used to
perform different operations. The ndarray (NumPy Array) is a multidimensional array used to
store values of same datatype. These arrays are indexed just like Sequences, starts with zero.
These arrays store the value of the distance between the eyes, nose, mouth and jaw, upper
outlines of the eye sockets, the sides of the mouth, location of the nose and eyes,
and the area surrounding the cheek bones.
OpenCV 2 uses machine learning algorithms to search for faces within a picture or video.
Because faces are so complicated, there isn’t one simple test that will tell you if it found a
face or not. Instead, there are thousands of small patterns and features that must be matched.
The algorithms break the task of identifying the face into thousands of smaller, bite-sized
tasks, each of which is easy to solve. These tasks are also called classifiers. For something

9
like a face, you might have 6,000 or more classifiers, all of which must match for a face to be
detected (within error limits, of course).

10
 Proposed Methodology

4.1 Proposed Algorithm

The software code of this system entirely relies on the architecture on the of the image
processing algrothims. The code, upon compilation start to get the input in the form the
camera footage and to generate output on the basis of the operations performed on the input
in respect with the trained dataset.

4.2 System Architecture & Flowchart

The flow chart of the system is shown below figure (a). The system algorithm comprises
of three main steps. First is to boot up the system, next step is to capture the faces that comes
in the CCTV Camera, this stage detects the face of person which is been seen by the camera.
It then sends the capture data to the image processing unit i.e. Raspberry PI where the faces
are matched by the stores faces in the database, if the faces are not matched with the faces
stored in the database then the system will send the text message to the registered number
informing that the intrusion has been detected. After that it will start the buzzer alarm.

Step 1: Power on the CCTV system.

Step 2: The CCTV camera start live detection face as soon as a person falls in its range where
it has been installed.
Step 3: Then the captured image will extract the faces and start matching with the faces store
in database.
Step 3.1: If the match will not found with the images stored in database, it will start
the security system commands and send a text message to registered mobile number
indicating that intrusion has been detected.
Step 3.2: The burglar alarm i.e buzzer will start ringing.
Step 4: Else
Step 5: Goto step 1.

11
 Start

CCTV Camera will detect the

faces comes in its range

Images capture by CCTV will

match stored faces in database

Is Yes
face
matched

No

Send a text message to

registered number & start
buzzer

Stop

Figure (a)

12
4.3 Software Implementation

The software design consists of a free running program which manipulates input capture
from the CCTV Camera and programming of the indicating unit, along with alarm unit. The
program code is written and develop on Microsoft VS Code and run on the virtual
environment which has been created using Anaconda Navigator.

13
 Future Scope And Limitations

This project can be used in in advance automatic cars in future to offer an advance
security system in automotive. It can be able to provide as a security module in advance
automatic cars in future like Tesla & MG.
It will consists of a face detection system, which detects the face of the unauthorised person
entered into the car, a web camera to capture the images and comparison is done using
MATLAB. The car engine will not start until the face of the driver doesn’t matches with the
stored faces in the database of car system.
If the face doesn’t match then the security system will send a text message to the owner of the
car informing that someone else than the registers driver face is trying to drive the car.

The limitations of this project are as follows:

• It can be only used in the isolated area where only limited persons are allowed to enter in
the CCTV surveillance zone.
• If the face has not been capture from a proper angle by the CCTV camera it may get
problems in matching the face within the database.
• The face only be recognised which are able to scan within the range of camera.
• If there is network issue then the text message is not going to send to the registered number.

14
 Conclusion

The innovation of this project will be very much beneficial to the people who are
looking for a cost effective face recognition CCTV security system for their security along
with a built GSM unit. The proposed idea of this project using embedded system will works
effectively. To reduce the errors in low light areas, we are going to used infrared CCTV
camera. A major advantage of the system is its low cost and ease in installation. Even after all
this, the system suffers from some disadvantages like problems in face detection majorly due
to appearance of people’s face from different angles. Apart from this, the system can be
considered an efficient one and can be used as a security measure for isolated areas.

15
 Bibliography

8.1 References

1. FindBiometrics, Facial recognition, [Online]

(Available at: http://findbiometrics.com/solutions/facial-recognition/)
2. Steve Mann, “Intelligent Image Processing”, Wiley-Interscience 2014.
3. https://www.w3schools.com
4. Walter G.Kropatsh, “Digital Image Analysis”, Springer 2013.
5. Wikipedia, [Online]
(Available at: http://en.wikipedia.org/wiki)
6. Bernd Jahne, “Digital Image Processing”, 5th Edition, Springer 2015.
7. https://realpython.com/face-recognition-with-python/

16