Sie sind auf Seite 1von 52

ABSTRACT

In the Entrance
Employee of each
subtleties andorganization, Receptionofis the
all differentInsubtleties there to check theThat
organization. Visitors subtleties
all are overseen and
by one
individual
Performs he is
crafted called Receptionist.
by visited
Receptionist this Project,
in a programmed we are creating
way.Algorithm
Here we use one framework that
Face acknowledgment to
discover
facial the Person
highlights ofinpeople. the organization.
In the event Open
thatthat
the the CV
newother
individual is
visitedutilized to separate
the organization the
implies
his face
implies is Stored
this framework Database.
book In the event
the Appointment old people visited the organization
methods it can
lessens the humandistinguish
work and at labor.
whatever point thefor that individual.
Person visited theWhen the faceThe
organization. is recorded
System
LIST OF FIGURES

Figure Page
no
Figure 1: Block Diagram 9

Figure 2: Circuit Diagram 10

Figure 3: Raspberry Pi 11

Figure 4: Raspberry Pi Pin diagram 12

Figure 5: Raspberry Pi Pin diagram(2) 12

Figure 6: Raspberry Pi camera 14

Figure 7: 2 channel relay 15

Figure 8: Schematic Diagram of the Relay 16

Figure 9: 2 channel relay with Raspberry Pi 17

Figure 10: DC Motor 18

Figure 11: LED 19

Figure 12: Installation of Raspbian 20

Figure 13: Installation of OpenCV 25

Figure 14: Edge features 27

Figure 15: Line Features 28

Figure 16: Center surround features 28

Figure 17: Special diagonal line features 29

Figure 18: Setup of the whole system for face recognition 30

Figure 19: Connection of the PC and Pi camera with Raspberry Pi 31

Figure 19(a): HDMI cable and Pi camera connection with Raspberry Pi 32


Figure 20: Output of Face Recognization 33

Figure 20(a): Output of the Face Rcognization 34

Figure 20(b): Outcome of the face recognization process 35


CONTENTS

Chapters Page No

Chapter 1: Introduction 1

Chapter 2: Literature Review

2.1: Earlier Developments 3


2.2: Recent Developments 4

2.3: Face Recognization 6

Chapter 3: Overview of the Project


3.1 Block Diagram 9

3.2 Hardware equipments 9

3.3 Circuit Diagram 10 10

Chapter 4: Component Description


4.1: Raspberry Pi 11

4.2 Raspberry Pi Camera 14

4.3 2 channel Relay 15

4.3.1 2 Channel Relay with Raspberry Pi 17

4.4 DC Motor 18

4.5 LED 19

Chapter 5: Methodology

5.1 Installing Raspbian on the Raspberry Pi 20

5.2 Installation of OpenCV 21

5.3 Face Recognization 24

5.4 Haar Cascade Classifier Algorithm 26


Chapter6: Results and Discussion 30

6.1 Face Recognization 33

Chapter 7: Conclusion and Future Works 37

Chapter 8: References 39

Chapter 9: Datasheets 42
INTRODUCTION

Home computerization implies the utilization of PC applications and data innovation for
control of home and private machines. Its application changes from essential controller of lighting to
complex microcontroller based frameworks including changing degrees of understanding and
robotization. Home computerization brings about simplicity of work, expanded productivity, and
security benefits provoking improved individual fulfilment. The pervasiveness of framework
engaged home robotization has been extending amazingly starting late due to straightforwardness
and much higher moderateness. Additionally, with the quick improvement of the Internet, there is
the potential for the controller and checking of such organization empowered apparatuses. In any
case, the new and empowering opportunities to extend the accessibility of devices inside the house
are the ultimate objective of home robotization. Displaying of a home robotization framework for
controlling all the different gadgets in Remote areas can be accomplished through an assortment of
correspondence alternatives, for example, wire less LAN advances, dial-up modems, satellite
correspondence, cell organization, Internet, etc.
The principle prudence of home computerization can be expressed as its security offices
where one can be advised and one can be frightened by this home robotization if there is any danger
caused to somebody's protection. Home computerization empowers people to exploit cutting edge
usefulness that was practically difficult to picture a couple of years back. The comfort factor here is
gigantic. Having the ability to keep most of the computerization inside home related through one
interface is an immense advance forward towards development and home mechanization.
Hypothetically, each of the one needs to do is basically sort out some way to use one application on
a mobile phone and tablet, and one will have the ability to exploit boundless limits and devices all
through the home. This diminishes the desire to ingest data for new customers, makes it more
straightforward to get to the handiness somebody truly requirement for their home. Brilliant home
frameworks will in general be magnificently adaptable with regards to the convenience of new
gadgets and apparatuses and other innovation. Regardless of how cutting edge your machines appear
today, there will be more up to date, more amazing models created over the long haul. Past that,
you'll most likely add to the set-up of gadgets as it will supplant the more established ones or find
new innovation to go with indoor and outside spaces.

6|Page
Having the option to coordinate these newcomers flawlessly will make everyones work as a
mortgage holder a lot simpler, and permits to continue moving up to the most recent way of life
innovation. Home computerization structures can relate movement locators, reconnaissance
cameras, robotized portal locks, and other generous security endeavors all through the home so one
can authorize them from one wireless before rushing toward bed. It will give security alerts on
various contraptions depending upon the hour of day a caution goes off, and watch exercises
persistently whether somebody is in the house or more than halfway around the globe. A significant
aspect of this security temperance of home robotization is keeping reconnaissance through cameras
and survilliance on the passage computerized and which can be just open by specific personalities
which are conceded consent to enter. For this doing confront recognization where in the event that
the face is perceived, at that point the entryway will open in any case not. In this proposed plot face
recognization which for the most part includes two phases which are face location and afterward
face recognization is finished.
Face Detection where the image is looked to take care of a face, at that point the image is
prepared to harvest and concentrate the individual's face for simpler recognization. Face
Recognization is where that distinguished and handled face is contrasted with an information base of
known countenances, to choose who that individual is. Here face location is performed on Intel's
open source structure called OpenCV. OpenCV (Open Source Computer Vision) is a library that can
be imported in practically all coding languages like python, C, Java and so forth python. Raspberry
pi-cam has been utilized to catch the image of the appearances that we will store in the information
base. Here in this proposed plot one will essentially experience three phases which are face
discovery, information gathering(where we will click image of the essences of the particular
characters) and the last stage will be the face recognization part which will coordinate the live faces
with the photos of the appearances in the data set and give us the ID subtleties of the given element
and work further for additional prerequisites. In the event that the coordinating file is half or more, at
that point no one but it very well may be said that it is a fruitful facial recognization measure.
OpenCV has got parcel of calculations however in this specific venture. Here Haar-Cascade
calculation Haar Cascade Algorithm has been utilized predominantly. This entire arrangement is
done on the working arrangement of Raspberry Pi which is known as Raspbian. Python has been
utilized here as the stage to play out the capacities and keeping OpenCV the primary space.

7|Page
At last home computerization can be summarized as the most helpful type of current
innovation. Home computerization is really a combination of a great deal of ethics meeting up like
security, control, comfort, convinience and in particular significant serenity. This structure is an
entirely adaptable type of innovation which is filling in the market quickly and is on the upsurge for
being a distinct advantage in the field of cutting edge computerization.

8|Page
CHAPTER 2: LITERATURE REVIEW

2.1PREVIOUS IMPROVEMENTS:
Home mechanization has been rotating around for around for quite a while. It's
exceptionally later that real keen homes have appeared. Maintaining as a top priority the points
of interest identified with home robotization innovative work extends as of now began from quite
a while back however because of the nonappearance of standard systems administration of the
home apparatuses no one could push ahead with this idea. The significant barricade was taken
out in 1992 with the improvement of the customer electronic transport by Electronics Industries
Association of America. The CEBus standard incorporates detail for a layered organization
design dependent on open framework interconnection model with network layer conventions for
the physical, Data Link, Network layer and application layer [1].
A proposed plot by Cross and Douligeris states that fiber optics may be the best mechanism
for home computerization network in light of the fact that with fiber, the impediments of the
home robotization framework can be decreased and capacities can be expanded to incorporate a
lot more capacities, prompting total home reconciliation. They saw that in spite of the fact that
CEBus incorporates fiber optics as one of the physical media however it isn't indicating the
arrangement of the fiber optic organization. After that they planned a fiber optics network which
offered different focal points, for example, (I) expanded transfer speed, (ii) invulnerability to
electromagnetic commotion, (iii) simplicity of establishment, and (iv) wellbeing from electric
stun perils. There were a couple of downsides too that the cost was high and as fiber optics can't
convey direct current they needed to utilize a substitute wellspring of energy [2-3]. In mid 1990s,
the purchaser hardware gadgets was developing into computerized design, hereafter the need was
felt to interconnect these home machines through advanced connects to safeguard the loyalty of
data sent. Chen proposed a home mechanization network which will fill the above need. Other
than the advanced connection fundamental component of the proposed home mechanization
network was the Digital access framework which because of which the home robotization
organization could convey and with the rest of the world moreover. Chen investigated the
utilization of IEEE 1394 for the proposed network as it could deal with both information and
isochronous traffic well at an information rate over 100 Mbps [4].

9|Page
Fujieda detected for flourishing and accomplishing total attractiveness. Home organizations
ought to be introduced effectively which will require network development without additional
wiring. Accordingly he proposed the plan for remote media in home organizations and named it
as remote home organizations. Fujieda proposed the utilization of 400MHz extraordinary low
power band for remote home organizations. He built up a little and low force microwave, climate
control system and clothes washers, Bluetooth passage and home terminal in a home. All
apparatuses had the Bluetooth unit inserted inside them and small associated with web [5].
In 2005 Hayong Oh and others, referenced the significance of energy effectiveness for the
directing plan of the sensors set at home. In the up and coming robotized home, sensors are
needed to be set wherever in the house which will gather different physical information, for
example, temperature, moistness, and light to give data to different machines. The creators talked
about that in the ordinary sensor directing plan each sensor hub recognizes an occasion and
afterward communicates the occasion to all sensor hubs inside one bounce range from where all
the hubs broadcast the message to the following hubs. This cycle is rehashed until the occasion
arrives at the base station. This plan for the most part cause seepage of battery power and the
batteries have restricted force so the steering plan needs to effective enough for fruitful usage of
home systems administration. From now on they prompted another sensor steering plan for home
mechanization organizations and called it as RDSR (Relative Direction based Sensor Routing).
This plan partitions the home zone into areas and finds a supervisor hub to every area. The sensor
gadgets move the information to director hubs in its area and afterward the information is moved
to the base station through the most brief way of two dimensional directions bringing about an
energy proficient plan [6]. In 2006

Mario Kolberg and Evan H.Magill referenced the control of complex arranged apparatuses.
By and by PC interface is generally used to design and distantly control these apparatuses.
Anyway the creators examined that this will be unsatisfactory for the intended interest group who
are not that much acquainted with the utilization of PCs. Accordingly they proposed Anoto-
empowered pen and paper as a reasonable option as clients are exceptionally acquainted with pen
and paper and they will think that its appropriate for control. In the proposed framework
information is moved to the specialist co-op through Bluetooth correspondence and versatile
handled and shipped off the client's home. Utilizing this methodology apparatus can be controlled
in the home and outside the home [7] 10 | P a g e
.2.2 RECENT DEVELOPMENTS:

Lately the Internet of Things (IOT) have empowered advancement to make day by day lives simpler
and agreeable. From coolers, to contraptions, to home security and numerous others can be controlled with
savvy home advancement. At the point when home gadgets are distantly controlled from anyplace regardless
of anything through web it turns into a significant ideals of the IoT (Web of Things). Security and living
greener are the fundamental highlights of the current homes. Current examples in home mechanization
incorporates distantly controlled machines and modernized devices, convenient/email/content alerts, and far
off video perception Sensors are the most significant aspect of this home computerization framework as it
goes about as the eyes of the framework. There are sensors for a broad assortment of employments, for
instance, estimating temperature, soddenness, light, liquid, and gas and perceiving advancement or uproar. If
there should arise an occurrence of Java Based Home Automation System the World Wide Web is utilized by
the home apparatuses to control and screen then the product motor is written in Java code that can be
convenient and furthermore can be utilized on any stage. The installed framework board is associated with
the pc home worker and serves the java pages and java beans [8].
If there should arise an occurrence of the Integrated Residential Gateway Controller for Home
Energy Management System, elite is inferred fro m conservative OS and to diminish the memory space than
the three stages plug and player component is utilized for decreasing establishment designing [9]. If there
should arise an occurrence of the Remote Controller for Home and Office Appliances by Telephone, the
locally established plan and the improvement of the telephone based distant regulator for the home and office
computerization, in this serious strategy the far off regulator is utilized for time and energy burning-through
[10]. The check of the proposed calculation, a few LED's were utilized to demonstrate the correct execution.
The Rpi has gotten an email through the client from a framework associated with the web. Gmail has been
utilized so as to direct mail correspondence. The relating results were created by a progression of email
correspondence. For instance an E-mail sent as "ON=1 ON Time=7:00. OFF Time=8:00 to the RPi account
from the client account. The calculation that is utilized beginnings its cycle by filtering the mail [11]. As per
R.A. Ramlee, keen home framework through Wireless Bluetooth is the will be the new goodness of this
mechanized advances where it is going about as an inspiration for impair people making it workable for them
to complete the every day movement, securely and easily [12].

11 | P a g e
Girish Birajdar and Shrikant Mahindrakar have built up a route in to computerize home apparatuses
utilizing an Embedded Web Server joined with a Raspberry Pi. Installed Web Server (EWS). is the installed
framework which can serve the web archives on demand from the customer. It by and large arrangements
with the board of dynamic substance and is anything but difficult to utilize and works at a quick speed. EWS
configuration incorporates a total web worker with TCP/IP uphold, running diverse OS, memory, application
savvy [13].

Android based home mechanization makes the framework more adaptable and gives alluring UI
contrasted with other home robotization frameworks. In this framework cell phones are incorporated into
home robotization frameworks. An epic design for a home computerization framework is proposed utilizing
the generally new correspondence advances. The framework comprises of chiefly three segments is a Wi-Fi
module, raspberry pi load up and hand-off circuits. Wi-Fi is utilized as the correspondence channel between
android telephone and the raspberry pi board [14].
2.3 FACE RECOGNIZATION:

Face location is one of the flow research points in the PC vision field. It is simple for people to
distinguish faces then again it is hard for PCs to identify faces.. The challenges related with face
discovery are varieties in scale, present, direction, lighting condition, outward appearance and so forth
Numerous methodologies have been actualized yet each has its own focal points and impediments. The
proposed framework is transfer on the appearance based methodology. Here the face identification is
finished by separating facial highlights like eye include, scaffold of the nose highlight, mouth highlight
and so forth which are available in a dark scale picture. These highlights are corresponding to the
adjustment conversely values between nearby gatherings of pixels however not to the force estimations
of a pixel. The highlights utilized in this framework are named as rectangular highlights and are
suggestive of Haar premise capacities [15] [16]. The proposed face location framework recognizes the
appearances with low bogus positive rate. At first the face discovery is finished with 10 and 12 phase
locator yet it gives helpless recognition precision and high bogus positive rate. The 14 phase indicator
gives great precision in face identification and extremely low bogus positive rates. The finder functions
admirably on Raspberry Pi with a 5MP camera and recognizes the essences of caught picture with the
least goal of 640x480 and the most elevated goal of 2560x1920pixels with adequate discovery speed
[17].

12 | P a g e
Patel and shah presented an exploration on facial element extraction methods for programmed
face explanation. An essential job is being played via programmed face comment part in sight and
sound data. The technique to recognize human countenances from picture and allot proper human name
is known as programmed face explanation. In face explanation face location and face acknowledgment
are the two most significant ideals. A creator likewise talk about the periods of the programmed face
comment and studied different methods of facial element extraction [18]. Patoliya and Desai created
ATM Security System utilizing Embedded Linux Platform which is being founded on the face location
framework. Raspberry pi is being utilized for the execution of this framework with the all-inclusive
capacities of Computer Vision (OpenCV) programming which is utilized for Image handling activity.
One prong secret word procedure is utilized in this framework which expands the security of the
framework. In an unapproved condition ATM entryway has been bolted and it's just when the OTP
secret word is entered my gatekeeper [19].
Heshmat et al. presented Face Identification framework in Video. A creator proposed CIE-Luv
shading space, facial element extraction and difference assessment. In observation, human machine
interfacing, Database the board framework can be actualized and is of incredible use as it is making an
appropriate visual interface here. The test results shows the productivity and the handiness of the
framework where the framework can perceive various faces and work appropriately and this one is
profoundly worthwhile because of it high security preferences [20]. Escalated research work and nitty
gritty examination on usage of savvy entryway lock framework utilizing GSM network by Jie-Ci Yang
[21]. Pi Camera Module is an extra with Raspberry Pi which goes about as a visual interface if there
should arise an occurrence of face location or Recognization, This interface utilizes the devoted CSI
interface, which was planned particularly for interfacing to cameras [22]. A calculation known as Haar
Cascade classifier is utilized for discovery of the item which it has been prepared for, from the source.
This proposed framework utilizes Haar Cascades classifier as a face location calculation [23]. Face
preprocessing is the module which diminishes the issues that makes the picture unclear to see the face.
for instance, less brilliance and unpredictability issues and commotion in the image and guarantee the
facial features constantly be in a reliable position. In this undertaking we use histogram leveling for
face preprocessing. For capability we use separate preprocessing which is histogram evening out for
left and right face. So histogram leveling is finished on different occasions, first thing for the whole
face and the other two for side appearances [24].

13 | P a g e
For better working, each phase of the course should have a low bogus negative rate, since, in
such a case that the real item is delegated a non-object, at that point the arrangement of that branch
stops, with no real way to address the mix-up made. Notwithstanding, each stage can have a generally
high bogus positive rate, on the grounds that regardless of whether the n-th stage arranges the non-
object as really being the article, at that point this slip-up can be fixed in n+1-the and resulting phases
of the classifier[25-26]. The reason or goal of the recommended work is to execute a working model of
a savvy entryway and to offer answers for the issue looked by individuals in everyday episodes of theft
or losing the key and furthermore to advance and touch off the work being done on IOT frameworks
and actualizing it with the assistance of key exploration regions of Neural Networks and IoT APIs and
protocols[27].

14 | P a g e
OVERVIEW
After the itemizedOFinvestigation
THE PROJECT: the CHAPTER
ofa Raspberry
past works, 3 in this proposed conspire building up a home
here
robotization
machines framework
and innovation utilizing
the most significant temperance Pi with
ofasthis which
task of onebewill
will the have
home the option to control
observation where home
the face
recognization
which is being utilized is which
OpenCV will be
whichutilized
is an open a piece
source PCsecurity
vision in this
libraryventure.
a piece The
of primary
python andstage
will
be utilized in Python.

CHAPTER 3: OUTLINE OF THE PROJECT

3.1 BLOCK DIAGRAM:

Miniature Door
Pi Camera LED

Raspberry Pi Relay DC Motor

Figure 1: Block Diagram

3.2 HARDWARE NECESSITIES:

 RASPBERRY Pi
 RASPBERRY Pi CAMERA
 4 CHANNEL 2 SLOT RELAY
 DC MOTOR
 LED
 BREADBOARD
 CONNECTING WIRES
 MINIATURE DOOR
3.3 CIRCUIT DIAGRAM:

Camera
slot Pi camera

Raspberry pi

GPIO DC Motor

Relay LED

Figure 2: Circuit diagram


Here one can see DESCRIPTION:
COMPONENT that the Pi camera is appended
CHAPTER 4 to the camera space and the DC engine and Relay is
joined to the GPIO pins of the Raspberry Pi. As it is a two channel transfer so it is associated with
the LED and the DC engine which helps in driving the DC motor.Here the DC engine drives a
smaller than usual entryway which is being utilized for an entryway locking framework for
reconnaissance purposes. The entryway just opens after an effective recognization measure, when
the face is put away already in the information base and that picture has the admittance to open the
entryway then it will work else it won't open.

CHAPTER 4: COMPONENT DESCRIPTION

4.1 RASPBERRY Pi:

Figure3:Raspberry Pi

In the above figure no.3 the Raspberry Pi is a low evaluated, little estimated PC that
connects to a PC screen or TV, and utilizations an ordinary peripherals like console and mouse. It is
a fit little gadget that empowers individuals, everything being equal, to investigate processing, and
to figure out how to program in dialects like Scratch and Python. It can do all that you'd anticipate
that a work station should do, from perusing the web and playing top quality video, to making
spreadsheets, word-preparing, and messing around. The Raspberry Pi was first dispatched in 2012,
and there have been a couple of changes and varieties performed starting now and into the
foreseeable future. The primary Pi had a solitary center 700MHz CPU and essentially 256MB
RAM, and the latest model has a quad-center 1.4GHz CPU with 1GB RAM.All over the world,
people use Raspberry Pi to get what it takes of programming capacities, , do home computerization,
and even use them in current applications.The Raspberry Pi works in the open source climate: it
runs Linux (different dissemination), and its guideline upheld working framework, Raspbian, is
COMPONENT
open DESCRIPTION:
source and runs CHAPTER
a set-up of open source4programming.

The Raspberry Pi Foundation adds to the Linux part and distinctive other open source
expands similarly as delivering its own one of a kind parcel programming as open source. Here
Raspberry pi is being utilized as a fundamental regulator to determine different highlights like face
recognition and identification which we are doing in our venture.

Figure 4: Pin diagram Raspberry Pi


COMPONENT DESCRIPTION: CHAPTER 4

Figure 5: Pin Diagram of Raspberry Pi(2)


An incredible element of the Raspberry Pi is the line of GPIO (universally useful info/yield)
pins along the top edge of the board. A 40-pin GPIO header is found on all current Raspberry Pi
sheets.

Voltages:
Two 5V pins and two 3V3 pins are available on the board, just as various ground pins (0V),
which are unconfigurable. The rest of the pins are altogether broadly useful 3V3 pins, which means
yields are set to 3V3 and inputs are 3V3-open minded.

Outputs:
A GPIO pin designated as an output pin can be set to high (3V3) or low
(0V).

Inputs:
A GPIO pin assigned as an information pin can be perused as high (3V3) or low (0V). This is
caused simpler with the utilization of inner draw to up or pull-down resistors. Pins GPIO2 and
GPIO3 have repaired pull resistors, however for different pins this can be designed in programming.
The GPIO pins can be utilized for different capacities too. Some are accessible on all pins and others
on some particular pins.

PWM (pulse-width modulation)


-Software PWM available on all pins
-Hardware PWM available on GPIO12, GPIO13, GPIO18, GPIO19
SPI
-SPI0: MOSI (GPIO10); MISO (GPIO9); SCLK (GPIO11); CE0 (GPIO8), CE1 (GPIO7)
-SPI1: MOSI (GPIO20); MISO (GPIO19); SCLK (GPIO21); CE0 (GPIO18); CE1 (GPIO17);
I2C
-Data: (GPIO2); Clock (GPIO3)
-EEPROM Data: (GPIO0); EEPROM Clock (GPIO)
Serial
-TX (GPIO14); RX (GPIO15)
4.2 RASPBERRY Pi CAMERA:

Figure 6: Raspberry Pi Camera


The Pi camera accompanies a flex link. The flex link is embedded into the connector which is
situated between the Ethernet and HDMI port with the silver connectors confronting the HDMI
port. The flex link connector is opened by pulling the tabs on the head of the connector upwards
then towards the Ethernet port. The flex link at that point is embedded solidly into the connector.
The top aspect of the connector at that point is pushed towards the HDMI connector and down,
while the flex link is held set up. Here the Pi camera is being used for the face identification and
the face recognization measure where right off the bat are caught and put away it in the
information base utilizing python and afterward again utilizing the camera while the
computerization and the observation part.

4.3 TWO CHANNEL RELAY:

Figure 7: 2 channel Relay

This is a LOW Level 5V 2-channel transfer interface board, and each channel needs a 15-
20mA driver momentum. It is utilized to control different apparatuses and gear with huge current.
It is outfitted with high-current transfers that work under AC250V 10A or DC30V 10A. It has a
standard interface that can be controlled straightforwardly by microcontroller. Transfer is being
utilized here as a driver circuit to drive different apparatuses through the hand-off.

FEATURES:
 Relay maximum output: DC 30V/10A, AC 250V/10A.
 2 Channel Relay Module with Optocoupler LOW Level Trigger expansion board,
which is perfect with arduino.
 Standard interface that can be controlled straightforwardly by microcontroller ( 8051,
AVR, *PIC, DSP, ARM, ARM, MSP430, TTL rationale)

Figure 8: Schematic diagram of 2 channel Relay

INPUT:
 VCC: Connected to positive supply voltage
 GND: Connected to negative supply voltage
 IN1: Signal triggering terminal 1 of relay module
 IN2: Signal triggering terminal 2 of relay module

OUTPUT:
 There is one NC(Normal-close), NO(Normal-open) and COM(common)
in each sub-modular of the relay so there are 2NC,2NO and 2COM.
4.3.1 RASBERRY Pi with TWO CHANNEL RELAY:

Figure 9: 2 channel Relay with Raspberry Pi

 Connect the signal terminal IN2、IN1 of 2-channel relay to port 17、18 of


the

Raspberry Pi, and connect an LED at the output terminal.

2-channel relay-----------------Raspberry Pi

IN2-----------------------17
IN1-----------------------18
4.4 DC MOTOR:

Figure 10: DC Motor


A DC engine is a class of turning electrical gear which converts direct flow electrical energy into
mechanical energy. The most well-known sorts rely upon the powers created by attractive fields.
Practically a wide range of DC engines have some interior instrument, either electromechanical or
electronic, which intermittently alters the course of current stream in part of the engine. A loop of
wire with a current going through it produces an electromagnetic field lined up with the focal point
of the curl. One can alter the course and the size of the attractive field by altering the course and
greatness of the current moving through it. In this proposed plot the DC engine is associated with the
transfer and drives the small scale entryway after fruitful recognization measure.

4.5 LED(Light Emitting Diode):

Figure 11: LED (Light Emitting Diode)


A light-emanating diode (LED) is a semiconductor light source that discharges light when current
courses through it. Electrons in the semiconductor recombine with electron openings, delivering
energy as photons. Here the light transmitting diode is utilized simply post face recognization where
it will show whether the face is perceived. In the event that the face is perceived the LED (Light
Emitting Diode) will sparkle else it won't gleam which will thusly open the entryway.

METHODOLOGY

5.1 Connecting Raspbian on the Raspberry Pi:


Introducing Raspbian on the Raspberry Pi is truly clear. Raspbian will be downloaded and
composing the circle picture to a miniature SD card, by then booting the Raspberry Pi to that
microSD card. For this endeavor, one needs a microSD card (with no under 8 GB), a PC with a
space for it, and, clearly, a Raspberry Pi and central peripherals (a mouse, support, screen, and
force source). This isn't the primary procedure for presenting Raspbian (more on that in a
moment), yet it's an important strategy to learn because it can similarly be used to present such a
critical number of other working undertakings on the Raspberry Pi. At the point when one
understands how to create a hover picture to a microSD card, we open up a lot of choices for
Raspberry Pi ventures.
Step1: Download the Raspbian
Turn on the PC and download the Raspbian circle picture. One can find the latest variation
of Raspbian on the Raspberry Pi Foundation's site here. It will require some investment,
especially in the occasion when one means to use the customary download elective instead of
the other download sources. It can absent a lot of take a stretch of half hour or more to
download.
Step2: Unzip the file
Turn on the PC and download the Raspbian plate picture. One can find the latest variation of
Raspbian on the Raspberry Pi Foundation's site here. It will require some investment, especially
in the occasion when one expects to use the traditional download elective rather than the other
download sources. It can absent a lot of take a stretch of half hour or more tThe Raspbian circle
picture is compacted, so it ought to be unfastened. The record utilizes the ZIP64 design, so
relying upon how current underlying utilities are, one needs to utilize certain projects to unfasten
it. Linux clients will utilize the fittingly named Unzip.o download.
Step3: Write the disc image to microSD card

Figure 12: Installing Raspbian on Raspberry Pi

One needs to pop the microSD card into our PC and compose the circle picture to it. The
cycle of really composing the picture will be somewhat unique over these projects, yet it's pretty
simple regardless of what is being utilized. Every one of these projects will have us select the
objective and the plate picture (the unfastened Raspbian record). Pick, twofold check, and afterward
catch to compose.
Step 4: Put the microSD card in your Pi and boot up
At the point when the circle picture has been stayed in contact with the microSD card, it is set
up to go. Put that microSD into your Raspberry Pi, plug in the peripherals and force source. The
current delivery to Raspbian will boot direct to the work area. Our default certifications are username
pi and secret phrase raspberry.

5.2 Installation of OpenCV:


Step1: The first thing to do is to expand our file system to include all available space on our
micro-SD card.
sudo raspi-config
sudo raspi-config

Step2 (install dependencies): The first step is to update and upgrade any existing packages.

sudo apt-get update && sudo apt-get upgrade.

Step3: Download the OpenCV source code


After having the dependencies installed we will now grab the 3.3.0 archive of OpenCV
from the official OpenCV repository

cd ~
wget -O opencv.zip
https://github.com/Itseez/opencv/archive/3.3.0.zip unzip opencv.zip

Step4: Python2.7 package manager

Before the start of compiling OpenCV on our Rasapberry Pi 3,one needs to install a python
package manager

wget https://bootstrap.pypa.io/get-
pip.py sudo python get-pip.py
sudo python3 get-pip.py

Step5: Creating Python virtual environment

Creating a Python virtual environment that will be used for computer vision development
Python2.7 is used.

mkvirtualenv cv -p python2

Step6: Checking whether we are in the “cv” virtual environment

In the event that Raspberry Pi is ever booted; log out and log back in; or open up another terminal
one needs to utilize the "workon" order to re-get to the "cv" virtual environment.

source ~/.profile
workon cv
Step7: Installing NumPy on our Raspberry Pi

The only Python dependency is NumPy, a Python package used for numerical processing:
pip install numpy

Step8: Complile and install OpenCV

OpenCV needs to be compiled and installed .First one should check whether one is in the virtual
environment then “execution of the command “workon” will be done.

workon cv

After being sure that one is in the “cv” virtual environment then one needs to setup the build
using CMake.

cd ~/opencv-3.3.0/
mkdir build
cd build
cmake -D CMAKE_BUILD_TYPE=RELEASE \
-D CMAKE_INSTALL_PREFIX=/usr/local \
-D INSTALL_PYTHON_EXAMPLES=ON \
-D OPENCV_EXTRA_MODULES_PATH=~/opencv_contrib-3.3.0/modules \
-D BUILD_EXAMPLES=O

Step9: We will now compile OpenCV

make -j4

sudo make install


sudo ldconfig

Step10: Finishing installing OpenCV on our Pi

ls -l /usr/local/lib/python2.7/site-packages/
total 1852
-rw-r--r-- 1 root staff 1895772 Mar 20 20:00 cv2.so

cd ~/.virtualenvs/cv/lib/python2.7/site-packages/
ln -s /usr/local/lib/python2.7/site-packages/cv2.so cv2.so

Step11: Testing our OpenCV 3.3.0

Open a new terminal, execute the source and workon commands, and then finally attempt
to import the Python + OpenCV bindings.

source ~/.profile
$ workon cv
$ python
>>> import cv2
>>> cv2. version
'3.3.0'
5.3 FACE RECOGNIZATION

To create a complete project on Face Recognition, we must work on 3 very distinct phases:

 Face Detection and Data Gathering

 Train the Recognizer

 Face Recognition

Step1: Instaling OpenCV 3.3.0

Figure 13: Installation of OpenCV


Step2: Checking Pi cam

After the establishment of open cv library, the above all else thing is to test if the pi camera is
working appropriately. On the off chance that the camera is appropriately introduced the to
progressing steps are to be performed. The camera must be adjusted in the right way to identify and
perceive the highlights of the face.
Step3: Face Detection and Data Gathering

Distinguishing faces and after that accumulate information. The most widely recognized approach to
distinguish faces is by utilizing Haar Cascade Algorithm. In this calculation continuous face
discovery is performed.
Step5: Trainer
On this subsequent stage, we should take all client information from our dataset and "coach" the
OpenCV Recognizer. This is done legitimately by a particular OpenCV work. The outcome will be a
.yml document that will be saved money on a "coach/" index.

Step5: Recognizer

Here catching a face will be done on the camera and if this individual had his face caught and
prepared previously, our recognizer will make a "forecast" restoring its id and a record, indicated
how sure the recognizer is with this match.
Here face recognization has been finished utilizing one sort of classifier which is known as Haar
Cascade classifier. In this cycle of face acknowledgment the most significant part is of picture
preparing. Picture handling is a mathematically genuine undertaking and probably the best region of
exploration for a significant data field. It is a methodology on the image to transform it into wanted
looking or which the information is an image and the yield may be an image or set of characters
related with the particular picture. It suggests assortment of procedures that are used to expand the
information yield from a picture.

5.4 HAAR-CASCADE CLASSIFIER ALGORITHM

It is essentially a classifier which is being utilized to recognize objects for which it has been
prepared for. Most importantly, the calculation needs a ton of positive pictures and negative
pictures to set up the Haar course classifier. Positive pictures will be pictures with clear faces where
negative pictures are those without any faces. Each component is spoken to as a solitary worth got
from the distinction of the entireties of pixels in white square shape from the whole of all pixels in
the dark square shape. All various potential sizes and areas of classifier is utilized for ascertaining
of a lot of highlights. As the quantity of classifiers increment the number-crunching calculations
appears to take quite a while. Rather than it, the idea of Integral Image has been utilized. Picture
Processing Integral picture is an information structure which is an added region table and
calculation for rapidly and effectively creating aggregate of qualities in a rectangular lattice subset.
To stay away from the intricate figuring we use Adaboost AI calculation, which is inbuilt in
OpenCV library that is course classifier, it takes out the repetition in the classifiers. Any classifier
which has a likelihood of half of additional in location is treated as frail classifier. The Sum of all
feeble classifier gives a solid classifier which settles on the choice about discovery. Order happens
in stages, if the chose locale falls flat in the main stage, we dispose of it. One doesn't utilize the
classifiers on that district which is disposed of. The locale which passes all the stages for example
all solid classifiers is treated as the distinguished face.

Identified Faces are passed to the Face acknowledgment stage. Nearby Binary Patterns histogram
calculation (LBPH) has been utilized for face acknowledgment. Nearby paired examples are basic
simultaneously extremely proficient surface administrator which relegates the pixels of the picture
by contrasting and the contiguous pixels as edge and which brings about a twofold outcome.
There are four types of features in Haar Cascade classifier algorithm:

 EDGE FEATURES
 LINE FEATURES
 CENTER-SURROUND FEATURES
 SPECIAL DIAGONAL LINE FEATURES

a) EDGE FEATURES:
Figure 14: Edge Features
b) LINE FEATURES:

Figure 15: Line Features

c) CENTER-SURROUND FEATURE:

Figure 16: Center-surround feature


d) SPECIAL DIAGONAL LINE FEATURES:

Figure 17: Special Diagonal Line Features


InRESULTS AND
this proposed DISCUSSIONS:
plot CHAPTER
it has been chosen 6 on a typical dc engine and associate it with a
to welcome
transfer which will be associated with the fundamental regulator Raspberry Pi. The DC Motor will
likewise associated with any of the pins of Raspberry Pi. DC engine here is helping the small way
to open after the face is perceived in Face Recognization measure. This entire framework is a post
part of the observation of Home mechanization.

RESULTS AND DISCUSSIONS

The principle goal of the proposed work here is to make a framework where it will be
anything but difficult to work home machines and supplies effectively by making this framework
easy to use. Building up a shrewd home framework was difficult from the start. The most
significant aspect of this proposed work is human reconnaissance which is significant because of
the security issues of keen homes. For reconnaissance utilizing face identification and face
recognization is being utilized which are the most present day type of observation. For this reason
Raspberry Pi is being utilized camera and OpenCV which is an open source which is a piece of
Python language. Python here goes about as the fundamental stage where the vast majority of the
work will be finished. Picture preparing should be accomplished for the Face acknowledgment.
OpenCV is an open source PC vision programming library. The library has a great deal of
upgraded calculations, which can be utilized in numerous IOT related areas including face
discovery and acknowledgment. As the libraries of our undertaking we got a kick out of the
chance to utilize the Haar classifier, LBPH (Lower Binary Pattern histogram) face recognizer.
Face recognization is should be fruitful if the coordinating record after recognization is over 45%.
RESULTS AND DISCUSSIONS: CHAPTER 6
Figure 18: Setup of the whole system for face recognization
In figure no 15 the entire framework is portrayed where one can see there is a case like structure
which is holding the Pi camera upstanding so it could take clear and legitimate pictures with the
end goal of face recognization.

Figure 19: Connection of the PC and Pi camera with Raspberry


Figure 19(a): HDMI cable and Pi camera connection with Raspberry Pi
In figure 16 and 16(a) it tends to be seen that how the PC, HDMI and the Pi camera is associated
with the principle regulator Raspberry Pi. The crate like structure is holding the Raspberry Pi
camera which is utilized for catching pictures of countenances to store in the information base and
afterward distinguish those appearances and perceive those countenances.

6.1 FACE RECOGNIZATION

Figure 20: Output of Face Recognization


The above figure is the result of the Face Recognization. This image is taken by the Raspberry Pi
camera which is utilized for the face recognization. First ly taking the photos and putting away it
in the information base then during the recognization cycle the result shows the name of the
character on the off chance that it is put away in the information base and it likewise shows the
coordinating record which is recently referenced that in the event that it is 45% or more, at that
point it is effective in any case not. Here unmistakably it isn't the situation as the coordinating file
is demonstrating 29.0%.

Figure 20(a): Output of the Face Rcognization


InCONCLUSION
the above figureAND
it is the yield of the
FUTRURE face recognization
WORKS: CHAPTER and 7 the individual is distinguished
appropriately. The coordinating record here is 44% which is practically 45% (the ideal coordinating
file) so one can say here that it is roughly fruitful.

Figure 20(b): Outcome of the face recognization process

In the above figure one can see that one face is being perceived as it is put away in the
information base and another face isn't appropriately perceived and is indicating "Obscure "from
here one can reason that on the off chance that the face isn't put away in the data set beforehand, at
that point the face won't be perceived and will show obscure
In this proposed plot it has been now chosen to present a keen entryway lock framework
where a scaled down entryway will be utilized. The entryway will be driven by a DC engine which
will be driven by a hand-off. the Pi camera is joined to the camera space and the DC engine and
Relay is connected to the GPIO pins of the Raspberry Pi. As it is a two channel transfer so it is
associated with the LED and the DC engine which helps in driving the DC engine .
CONCLUSION AND FUTRURE WORKS: CHAPTER 7

Here the DC engine drives a small entryway which is being utilized for an entryway locking
framework for observation purposes. The entryway just opens after a fruitful recognization
measure, when the face is put away already in the information base and that picture has the
admittance to open the entryway then it will work else it won't open.
CONCLUSION AND FUTRURE WORKS: CHAPTER 7

CONCLUSION AND FUTURE WORKS

Home Automation is the most moving innovation as of now. This type of innovation
incorporates a great deal of possibilities from a ton of themes or subject gathered together. Home
mechanization here has nearly gotten a developmental change taking care of or working home
machines and supplies and made it simple and helpful to utilize home apparatuses. A ton of home
machines can be controlled through home robotization, for example, lights, fans, TV, forced air
systems, cooler and so on Home robotization encourages us in working this machines from far
away. In this proposed conspire an exertion has been made to assemble a home computerization
framework with the principle righteousness of Face Recognization. Face Recognization here has
been finished with the assistance of a Raspberry Pi camera which was utilized to take photos of
the countenances and store it in the information base where. The entire recognizarion has been
done on Open CV which is a library work in Python language and is an open source. OpenCV is
an open source PC vision programming library. The library has a great deal of upgraded
calculations, which can be utilized in numerous IOT related areas including face identification
and acknowledgment.
As the libraries of our undertaking we got a kick out of the chance to utilize the Haar
classifier, LBPH (Lower Binary Pattern histogram) face recognizer. Face recognization is should
be fruitful if the coordinating file after recognization is over 45%. In figure20 we are seeing the
yield of the Face Recognization. Right off the bat taking the photos and putting away it in the
information base then during the recognization cycle the result shows the name of the character in
the event that it is put away in the information base and it additionally shows the coordinating list
which is recently referenced that in the event that it is 45% or more, at that point it is fruitful in
any case not. Here obviously it isn't the situation as the coordinating file is indicating 29.0%. In
figure 20(a) one can see that the yield of the face recognization and the individual is
distinguished appropriately. The coordinating record here is 44% which is practically 45% (the
ideal coordinating file) so we can say here that it is roughly fruitful. Also, in another image one
other face has been included where the picture of the face was not put away already so
accordingly it is demonstrating obscure. In this proposed conspire an entryway lock framework is
likewise going to be utilized which will help in making the security part more solid.
CONCLUSION AND FUTRURE WORKS: CHAPTER 7
REFERENCES:
The entryway will CHAPTER
be driven by8a DC engine which will be associated with a transfer. To better the
Face acknowledgment execution, there are different things that can be improved here, some of them
being genuinely simple to go with. For instance, you could include shading preparing, edge
identification, and so forth Today, one of the fields that uses facial acknowledgment the most is
security. Facial acknowledgment is an incredibly suitable contraption that can help law aces see
guilty parties and programming associations are using the advancement to empower customers to get
to their development. This development can be moreover made to be used in various streets, for
instance, ATMs, getting to mystery archives, or other delicate materials. This can put forth other
security attempts, for instance, passwords and keys old. Another way that innovators are planning to
execute facial affirmation is inside cable cars and other transportation outlets. They are planning to
utilize this advancement to use faces as charge cards to pay for your transportation cost. As opposed
to embarking to a slow down to buy a ticket for a confirmation, the face affirmation would take your
face, run it through a structure, and charge the record that you've as of late made. This could
possibly smooth out the strategy and advance the flood of traffic unquestionably.

[1] Electronics Industries Association, Consumer Electronics Group, EIA Home Automation
System: EIA/Is-60, EIA ,1992.

[2] D.L. Cross, C. Douligeris; “A Fiber Optic Home Automation System”, IEEE Transactions on
Consumer Electronics, Vol. 39, No. 3, August 1993,pp. 636-645.

[3] C. Douligiris; “Intelligent Home Systems” IEEE Communication Magazine, October 1993,
pp. 52-61

[4] W.Y. Chen; “Emerging Home Digital Networking Needs”, Proc. 4th IEEE Int. workshop on
Community Networking, Atlanta, 1997, pp. 7-12..

[5] H. Fujieda, Y. Horiike, T. Yamamoto and T. Nomura; “A Wireless Home Network and its
Application Systems”, IEEE Transactions on Consumer Electronics, Vol. 46, No. 2, May 2000,
pp. 283-290.

[6] H.Oh, H.Bahn, K.J. Chae; “An Energy-Efficient Sensor Routing Scheme for Home
Automation Networks”, IEEE Transactions on Consumer Electronics, Vol. 51, No.3, August
2005.

[7] Mario Kolberg & Evan H. Magill; “Using Pen and Paper to control Networked Appliances”,
IEEE Communications Magazine, November 2006, pp. 148-154.

[8] Al-Ali A. R and Al-Rousan M., "Java-based home automation system", IEEE
Transactions on Consumer Electronics, vol. 50, 2003,no. 2, pp. 498-504.

[9] N.Kushiro., S.Suzuki., M.Nakata., H.Takahara and M.Inoue,” Integrated


Residential Gateway Controller For Home Automation System” IEEE Transactions On
Consumer Electronics, vol. 59, No. 3 August 2003.
REFERENCES: CHAPTER 8
[10] I.Cogkun and H.A.G.Universitesi, T.E.Fakultesi, E.E.Bolumu., “A Remote Controller For
Home And Office Appliances By Telephone”, IEEE Transactions on Consumer Electronics, Vol.
44, No. 4, November 1998.

[11] Narendra M and Vijaylakshmi M,"Raspberry pi based advanced scheduled


home automation system through E-mail", Chennai, 2014.
[12] R.A.Ramlee, D.H.Z.Tang dan M.M.Ismail(2012), “Smart Home System For Disabled
People Via Wireless Bluetooth”, IEEE International Conference on System Engineering
and Technology ,pp. 978-983,2012.

[13] G.Birajda and S.Mahindrakar, “Embedded web server based home automation using
raspberry pi”, International Journal of Modern Trends in Engineering and Research, vol. 1, no.5
,,India,September2014.

[14] H Bharathi ,U Srivani, MD Azharudhin, M Srikanth and M Sukumarline, "Home


Automation by Using Raspberry Pi And Android Application", Department of Electronic s
and Communication Engineering, MLR Institute of Technology, Hyderabad,India,2017.

[15] Paul Viola and Michael Jones, “Rapid object detection using boosted cascade of simple
Features”, IEEE Conference on Computer vision and Pattern Recognition, 2001, Vol.1,pp. 511-
518.

[16] C. Papageorgiou, M. Oren and T. Poggio, “A general framework for object detection”,
InInternational Conference on Computer Wsion, Bombay, Jan 1998, pp. 555-562.

[17] N.V. Naik, A.Venunadan and Kumara .K .R,"Face Detection on Open CV using Raspberry
Pi", Department of ECE,GSIT, Karwar, Karnataka, IJEECS, volume6, 2017.

[18] T.Patel and B.Shah, “A Survey on Facial Feature Extraction Techniques for Automatic Face
Annotation”, IEEE International Conference on Innovative Mechanisms for Industry
Applications (ICIMIA),Bangalore, India, 2017.

[19] J.J. Patoliya and M.M. Desai, “Face Detection based ATM Security System using
Embedded Linux Platform”, IEEE 2017 2nd International Conference for Convergence
in Technology (I2CT),Mumbai, India,2017.

[20] M.Heshmat, W.M. Abd-Elhafiez, M.Girgisand S.Elaw, “Face Identification System


in Video”, IEEE 2016 11th International Conference on Computer Engineering &
Systems (ICCES), Cairo, Egypt,2016.
[21] J.C Yang, "An Intelligent Automated Door Control System Based on a Smart
Camera".2012.

[22] W. F. Abaya, J. Basa, M. Sy, A. C. Abad and E.P. Dadios, "Low cost smart security
camera with night vision capability using Raspberry Pi and OpenCV," 2014 International
Conference on Humanoid, Nanotechnology, Information Technology, Communication and
Control, Environment and Management (HNICEM), Palawan, 2014, pp. 1-6

[23] P.Pasumarti1, P. P. Sekhar “Classroom Attendance Using Face Detection and Raspberry-
Pi”
International Research journal of Engineering and Technology (IRJET), Volume: 05
Issue: 03. p3-p5,Mar-2018

[24] S.Rajkumar, J.Prakash, “Automated attendance using Raspberry pi”, International


Journal
Of Pharmacy & Technology (IJPT), Vol. 8, No. 3, pp. 16214-16221, September 2016.-

[25] T. M. Inc., “Train a Cascade Object Detector,”[Online]. Available:


http://www.mathworks.se/help/visio n/ug/traina-cascadeobject -d

[26] S.Roy, Md.N.Uddin, Md.Z.Haque,Md.J Kabir,"Design and Implementation of the


Smart Door Lock System with Face Recognition Method using the Linux Platform
Raspberry Pi",IJCSN - International Journal of Computer Science and Network, Volume
7, Issue 6, December 2018.

[27] N.Saini, S.Kaur and H.Singh A Review: Face Detection Methods And Algorithms,
International Journal of Engineering Research & Technology (IJERT) ISSN:2278-0181
www.ijert.org Vol. 2 Issue 6, June – 2013.

Das könnte Ihnen auch gefallen