International Journal of Advances in Engineering and Management (IJAEM)

Volume 2, Issue 10, pp: 63-66 www.ijaem.net ISSN: 2395-5252

An IoT device: Smart Door Bolt for home security using

Face Recognition
Rohini Janardan Auti1
1
Academic Research Student, Department of Information Technology B. K. Birla College of Arts, Scienceand
Commerce (Autonomous), Kalyan

-------------------------------------------------------------------------------------------------------------------------------------
Date of Submission: 20-11-2020 Date of Acceptance: 03-12-2020
--------------------------------------------------------------------------------------------------------------------------------------
face recognition has the most effective and
ABSTRACT: Our door is the first and last thing valuable advantage that’s it hassle-free. In a sense,
which come across while entering and leaving the that individual does not need to touch anything to
home, that’s why the home automation in the door get identified or recognized. This is the reason for
lock system enables users more conveniently to the growth of biometric devices of face-
control and monitor the home environment. This recognizing.
paper focuses on the prime aspect ofsecurity and Face recognition is a process which first
easy access of smart door lock can provide rather involves face detection, then face alignment,
than traditional method of door locking system. feature extraction, and at last face recognition. In
Most people have an idea about digital lockers but face detection process one or more faces get
smart door bolts are one step ahead of it. Having a located in the image and mark with a bounding
lock and key for the door are the issues in the box. At the stage of face alignment system
Traditional method of home security, it is easily normalize theface to be consistent with the
breakable and quite outdated too. that’s a result of a database, such as geometry and photometrics. Then
robbery. To tackle this problem, the concept has feature extraction extract features from the face that
intelligent door pop up. In Smart door bolt, there can be used for the recognition task, and at the last
aremany methods are used to unlock the door like stage facerecognition perform matching of the face
smart card scanner, biometric (finger, palm, DNA, against the known faces in a prepared database.
iris, Face), specific code, etc. from all this the face Though at the time of the installation of the device,
recognition is one which gives access with the door the faces of family members get scanned and store
without touching anything and also gives more as a fix data. After that whenever the object or
safety. Using face recognition in a security system person comes in front of the door the motion-
will give good results in multi-face detection and triggered camera captures the pictures and the real-
stranger identification which are the most prior time face detection program gets starts by using the
requirements that come under our home security. Local Binary Pattern (LBP) method. And then if
Keywords:Smart door bolt, security, Face the person's image matches with anyone from the
recognition, motion camera, conveniency, IoT. family member the door will immediately get
unlock. If the face is not identified i.e. stranger’s
I. INTRODUCTION face comes in front of the door then the doorbell
The Security of our home is a crucial issue will ring. And the most important that’s if someone
nowadays. Everyone is looking for a safe, tries to break or damage the door the alarm will be
protected, and trustable device for our door locking raised. As a result, we can surely say that face
system. Everyone wants that our home should be recognition is one of the best security systems.
safe and protected for us and our loved ones.
keeping this in mind many of the countries II. OBJECTIVES
developing new ideas and devices for home A. To study if a Smart door bolt security system
security systems day by day. From the past few using face recognition provides more
days, it is important to have a safe, convenient, and protection and safety measures as compared to
ideal security system for door lock. Instead of using a regular door bolt.
the card, password, key, pattern for lock using the B. To study if a Smart door bolt security system
biometric data is safer and also ideal too, because it using face recognition provides more
does the automatic person identification from the protection and safety measures as compared to
inputs such as fingerprint scanner, palm scanner, a regular door bolt.
DNA analyzers, face recognition. From all this,
DOI: 10.35629/5252-02106366 | Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 63
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 2, Issue 10, pp: 63-66 www.ijaem.net ISSN: 2395-5252

C. To analyse how Face recognition is more also gives the warning to the stranger via LED
preferable among other biometric devices. display by using real-time face recognition and face
detection. In [10], the authors proved that how the
By using the survey analysis, the following IoT is a need of the era in the field of technology of
hypothesis is proposed to attain the above smart doors. In the proposed system they the
objectives: exposed smart door lock in the integration of sets
H1: Individual using smart door lock experience of devices were keeping the IoT as a backbone of
features like: (a). Security, (b). Easy to use, (c). the system and face recognition as one of the tools
cost-efficiency. of a respective security system. In [15] also the
system proposed by the author provides a
III. RELATED WORK convenient way to control and monitor the door via
In [1], the article proposed by the authors mobile application.
explained how the wireless system provides more
flexibility and extensibility than the wired one in- IV. METHODOLOGY
home automation security system. keeping the Using an online survey is held which
ZigBee module as a backbone of a digital wireless contains the questionnaires to test the proposed
door lock system which exploits the full capacity of hypothesis. Link of that form was circulated
the ZigBee sensor network by integrating home through social media platforms to my target
security by home automation. There are fewer population. From the responses, certain parameters
cases reported by the media of digital door locks get concluded to obtain the solutionof objectives.
are opened by some invalid users to invade homes
and offices than regular door bolt, but cases are A. PARTICIPANTS:
there is more focusable point and keeping this in Based on two conditions i.e. security and
the mind Ha, I. [2] proposed a system which convenience the proposed hypothesis gets tested.
provides strengthened security functions that can A total of 55 respondent data was collected from
transfer the recorded images to the user’s mobile the target population. Among the 55
device when an invalid user/ stranger attempt any population,55% were male and 45% were female.
illegal operation. This system makes the user
enable to remotely access the door lock and B. MEASURES:
increases the convenience to operate it. In paper Gender Security Conveniency Total
[3], the authors are focused mainly on the design of
the system which is cheap, simple, can be installed Male 18 12 30
easily, and also can be fully customized based on
application specific requirements. a real-time live
streaming and monitoring system using Raspberry Female 10 15 25
Pi with installed wi-fi connectivity has explained
the model which is proposed in [4] by Solanki, V. Total 28 27 55
P., & Deshmukh, S. To overcome the problems like
robbery and identity fraud how the face recognition Observed Value
using deep learning and IoT techniques is used in a
door lock system is explained in [5], also this The above-observed table shows the data
article concludes that combination of face obtained by the survey form and for calculate the
recognizer with IoT experimentally gives a expected data there is a simple formula:
successful result. By having the combination of Expected value= (row total) *(column
microcontroller esp8266 and smartphone with the total) / (Grand total)
respective application the system gets proposed in
paper [7] which makes the door intelligent. This
system makes the user control the locking and
unlocking of the door via a mobile application. By
using the IoT and face recognition Pawar, S.,
Kithani, V., Ahuja, S., & Sahu, S. proposed a
system in [8], which deliver a cost and energy-
efficient solution for home security. In that, they
have explained how the device can send a Expected Value
notification to the homeowner via SMS and email
if any stranger gets detected in front of the door and

DOI: 10.35629/5252-02106366 | Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 64
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 2, Issue 10, pp: 63-66 www.ijaem.net ISSN: 2395-5252

From observed value and expected value, we can anyone who knows unlocking face he/she could
apply the formula of the chi-square test unlock the door just using a photograph of that
2
Oi – Ei person. In the future, while implementation face
X2=∑ where, Oi= Observed value, recognition into the smart door lock there should be
Ei
Ei= Expected a restriction to allow only verified software that can
value detect 3D and moving objects only.

V. EXPERIMENT VIII. CONCLUTION

After collecting the observed and expected When it comes to accessibility, security,
values test score of independent samples was convenience, face detection in a smart door lock is
calculated at the significance level of 95% using one of the best security systems. But still only in
the chi-square test from the formula of chi-square the huge towers, in high society area, this IoT
X2 is obtained i.e. 2.186909927. the tabulated value device gets installed. By increasing the productivity
for chi-square is 3.84 and after calculation and availability of smart door bolts in the market
2.186909927 is the value obtained as a calculated with some decrement at the starting in the cost a
value. So, plotting both the value on a graph, common man can also prefer to use it.The study in
statistics of Chi-square test formed with the degree this paper, reaches till the prime aspect of it.
of freedom 1.
ACKNOWLEDGEMENT
The author would like to present special
gratitude to Prof. Swapna Nikale, Department of
Information Technology of B.K. Birla college of
Arts, Science and Commerce (Autonomous)
Kalyan, Thane and to all the respondent for their
valuable time and fruitful data by survey form.

REFERENCE
[1]. Park, Y. T., Sthapit, P., & Pyun, J. (2009).
Smart digital door lock for the home
automation. TENCON 2009 - 2009 IEEE
Region 10
Conference.https://ieeexplore.ieee.org/docu
ment/5396038
Fig1. STATISTIC OF CHI-SQUARE TEST [2]. Ha, I. (2015). Security and Usability
Improvement on a Digital Door Lock
VI. RESULT System based on Internet of Things.
The score of independent data using the International Journal of Security and Its
survey analysis and calculated Chi-square value by Applications, 9(8), 45-
the experiment resulted that participants experience 54.http://article.nadiapub.com/IJSIA/vol9_n
better security and convenience in an IOT based o8/5.pdf
smart door bolt system using face recognition. The [3]. Nath, S., Banerjee, P., Biswas, R. N., Mitra,
proposed device can fulfill and satisfies all the prior S. K., & Naskar, M. K. (2016). Arduino
requirements of the user. Hence, “H1” is accepted. based door unlocking system with real time
control. 2016 2nd International Conference
VII. LIMITATIONS AND FUTURE on Contemporary Computing and
SCOPE Informatics
This research study utilized a sample of (IC3I).https://ieeexplore.ieee.org/document/
the target population who were possessive about 7917989
respective home security. The study only included [4]. Solanki, V. P., & Deshmukh, S. (2019). Wi-
limited parameters that can achieve the objectives Fi Based Home Surveillance Bot. 2019 5th
but not enough to complete the implementations International Conference On Computing,
happening in the area of security systems. Communication, Control And Automation
Implementation of face recognition in smart door (ICCUBEA).
lock obviously gives secure accessibility but fraud https://ieeexplore.ieee.org/document/912935
can happen with some cheap devices. Not using the 8
proper software leads to detect 2D images also. So,
DOI: 10.35629/5252-02106366 | Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 65
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 2, Issue 10, pp: 63-66 www.ijaem.net ISSN: 2395-5252

[5]. Radzi, S. A., Alif, M. K. M. F., Athirah, Y. and Emerging Technology (ICSCET).
N., Jaafar, A. S., Norihan, A. H., & Saleha, doi:10.1109/icscet.2018.8537344
M. S. (2020). IoT based facial recognition [11]. Arjona, A. B., Bautista, P. M., Edma, J. E.,
door access control home security system Martel, M. P., & Octavio, E. N. (october
using raspberry pi. International Journal of 2019). Design and Implementation of an
Power Electronics and Drive Systems Arduino-Based Security System Using Laser
(IJPEDS), 11(1), 417. Light. Arianne B. Arjona. LPU-Laguna
http://ijpeds.iaescore.com/index.php/IJPEDS Journal of Engineering and Computer
/article/view/20490 Studies, Vol. 4(No.
[6]. Chen, J., Li, S., Zhang, Y., & Raychaudhuri, 2).https://lpulaguna.edu.ph/wp-
D. (2017). Motion-Triggered Surveillance content/uploads/2019/10/2.-Design-and-
Camera using MF-IoT. Proceedings of the Implementation-of-an-Arduino-Based-
Second International Conference on Security-System-Using-Laser-Light.pdf
Internet-of-Things Design and [12]. Rai, A., Rai, M., Jogi, N., Rai, B., Rai, S., &
Implementation. Rasaily, D. (2019). Low Cost Laser Light
https://dl.acm.org/doi/10.1145/3054977.305 Security System in Smart Home. 2019
7291 International Conference on Innovative
[7]. Ribeiro, V. P., Oliveira, L. da S., Sustainable Computational Technologies
Nascimento, D. A. do, Alencar, D. B. de, & (CISCT).
Júnior, J. de A. B. (2019). Application of the doi:10.1109/cisct46613.2019.9008141
Internet of Things in the Development of a [13]. Davda, P. A., Ahirrao, S. B., & Dangwani,
“Smart” Door. International Journal of A. (2020). IoT Based Smart Door Lock
Advanced Engineering Research and System. Journal of Information and
Science, 6(5). Retrieved from http://journal- Computational Science, Volume 13(Issue 6
repository.com/index.php/ijaers/article/view/ – 2020).
824 https://thesai.org/Publications/ViewPaper?V
[8]. Pawar, S., Kithani, V., Ahuja, S., & Sahu, S. olume=10&Issue=5&Code=IJACSA&Serial
(2018). Smart Home Security Using IoT and No=56
Face Recognition. 2018 Fourth International [14]. Anvekar, R. G., & Banakar, R. M. (2017).
Conference on Computing Communication IOT application development: Home
Control and Automation security system. 2017 IEEE Technological
(ICCUBEA).https://ieeexplore.ieee.org/docu Innovations in ICT for Agriculture and Rural
ment/8697695 Development
[9]. Agarwal, A., Mehandiratta, E., Sanket, R., (TIAR).https://ieeexplore.ieee.org/document
Samkaria, R., Gupta, T., Singh, R., & /8273688
Gehlot, A. (2016). Smart Door Lock System [15]. Asnani, H., Khan, S., Nandeesh, S., & T.V,
for Elderly, Handicapped People Living P. (aug 2018). Securing an IoT based Home
Alone. International Journal of Smart Home, using Digital Image Processing and an
10(6), 155–162. Android Application. International Research
https://gvpress.com/journals/IJSH/vol10_no Journal of Engineering and Technology
6/16.pdf (IRJET), Volume: 05(Issue: 08), 410-
[10]. Balla, P. B., & Jadhao, K. T. (2018). IoT 415.https://www.irjet.net/archives/V5/i8/IRJ
Based Facial Recognition Security System. ET-V5I872.pdf
2018 International Conference on Smart City

DOI: 10.35629/5252-02106366 | Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 66