See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/317102430

Implementation of IoT-Based Smart Video Surveillance System

Conference Paper  in  Advances in Intelligent Systems and Computing · May 2017

DOI: 10.1007/978-981-10-3874-7_73

CITATIONS READS

9 6,458

3 authors, including:

Suchitra Khoje Prajakta Pardeshi

Dr. Vishwanath Karad MIT World Peace University, Pune. College of Engineering, Pune
23 PUBLICATIONS   89 CITATIONS    4 PUBLICATIONS   13 CITATIONS   

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Suchitra Khoje on 09 April 2018.

The user has requested enhancement of the downloaded file.

Implementation of IoT-Based Smart
Video Surveillance System

Sonali P. Gulve, Suchitra A. Khoje and Prajakta Pardeshi

Abstract Smart video surveillance is a IOT-based application as it uses Internet for

various purposes. The proposed system intimates about the presence of any person
in the premises, also providing more security by recording the activity of that
person. While leaving the premises, user activates the system by entering password.
System working starts with detection of motion refining to human detection fol-
lowed by counting human in the room and human presence also gets notified to
neighbor by turning on alarm. In addition, notification about the same is send to
user through SMS and e-mail. The proposed system’s hardware implementation is
supported by Raspberry Pi and Arduino board; on the other hand, software is given
by OpenCV (for video surveillance) and GSM module (for SMS alert and e-mail
notification). Apart from security aspect, system is intelligent enough to optimize
power consumption wastage if user forgets to switch off any electronic appliances
by customizing coding with specific appliances.

Keyword IOT (Internet of Things)

1 Introduction

In the present world, situation security assumes a vital part. Numerous individuals
utilize distinctive sorts of security system to keep their property from unapproved
person’s entry. Security system helps individuals to feel somewhat safe while they
have to travel or avoid their home for work. A large number of the security system
works just inside a specific territory limit [1], for instance, CCTV, as a person need

S.P. Gulve (✉) ⋅ S.A. Khoje ⋅ P. Pardeshi

MAEER’s MITCOE, Kothrud, Pune (MH), India
e-mail: sonaligulve@gmail.com
S.A. Khoje
e-mail: suchitra.khoje@mitcoe.edu.in
P. Pardeshi
e-mail: prajakta.pardeshi@mitcoe.edu.in

© Springer Nature Singapore Pte Ltd. 2017 771

H.S. Behera and D.P. Mohapatra (eds.), Computational Intelligence
in Data Mining, Advances in Intelligent Systems and Computing 556,
DOI 10.1007/978-981-10-3874-7_73
772 S.P. Gulve et al.

to see camera footage from control room. The current security systems against
robbery are entirely costly as a certain measure of cash must be paid to adminis-
tration supplier to store the recorded video despite the fact that there is no human
movement is recognized. The solution for this problem is an intelligent surveillance
system that can start recording video only after a human motion is detected. This
eventually minimizes the required storage space and makes system cost-effective.
The proposed framework gives more security with the assistance of Web at less
expensive cost and requires less storage space. In literature [2 and 3], researchers
have proposed various methods for people counting. In literature [4–10],
researchers have proposed many image processing methods/algorithms for human
counting which are prone to problems such as occlusion or shadow and overlap-
ping. To address these problems at some extent, Rossi and Bozzoli [4] and Sexton
et al. [5] proposed a technique in which the position of camera is vertical as for the
plane of the floor. In literature [11], researchers proposed an improved adaptive
background mixture model for real-time tracking with shadow detection. The
proposed framework gives a smart security system which gives home security with
SMS and e-mail notice about the unapproved people nearness, programmed human
checking and switching off all the appliances which consumes more power by
customizing coding with particular appliances. Proposed system performs various
tasks such as motion detection, human detection and counting, alarm activation,
SMS notification through GSM and Internet Twilio account, and e-mail
notification.
To improve the system performance, two boards are used—Raspberry Pi and
Arduino. Raspberry Pi works in surveillance mode and Arduino works in normal
mode. Arduino verifies the password and allows Raspberry Pi to start the surveil-
lance mode. Once the password is verified, Arduino turns off all the electrical
appliances by customizing coding with specific appliances. Raspberry Pi performs
various tasks in surveillance mode such as motion and human detection, human
counting, sending SMS, and e-mail notification to user after human detection. After
human detection, Raspberry Pi sends command to Arduino for sending SMS to user
by communicating with GSM module. By default, system remains in normal mode.
As the user enters correct password, system starts working in surveillance mode.
In surveillance mode, Raspberry Pi detects human motion and counts number of
people in a room. The location of a camera is at the entrance of a room. The human
count is implemented by background subtraction [2] method in OpenCV. If any
human is detected in surveillance mode, then using the GSM module and Twilio
account message is sent to the owner of the house. The highlights of proposed
system are as follows:
(1) The proposed framework includes people counting, and two notices are sent to
client by SMS: One SMS is sent through GSM and one SMS is sent through
Twilio trial account with the assistance of Web. The recorded video is sent as a
e-mail to client. At the point when there is no individual in the premises, the
framework works in ordinary mode.
Implementation of IoT-Based Smart Video Surveillance System 773

(2) Raspberry Pi detects motion and human presence and it counts number of
humans in a room. As the system detects human presence, immediately a SMS
notification is sent to the user. The system also sends the recorded video to
users mail id. As a human is detected, GSM module gets instruction from
Arduino regarding SMS notification. Another SMS notification is sent through
Internet Twilio trial account. The alarm is turned on as human presence is
detected.
(3) The proposed system also provides a facility to control electrical appliances by
turning them off. The proposed system offers few advantages such as-
(i) Less memory storage space is used for recording video as system start
recording the video only after motion is detected.
(ii) Recorded video is e-mail to user so that the user can inspect it later.
(iii) User gets noticed (SMS and Email) just after human detection, so that he
can take necessary actions immediately.

2 Working Principle

The proposed framework is initiated by entering right password. The movement

recognition algorithm is actualized to distinguish the moving items and human
count in room is done by utilizing OpenCV. After the password verification system
starts working in surveillance mode and all the electrical appliances are turned off
appliances by customizing coding with specific appliances. If motion is detected,
the system checks for human detection. As system detects human presence, a
notification through SMS is sent and alarm is turned on. The activity of that human
is recorded and e-mailed to user.
If the video consists of less than or equal to 100 moving frames, the video is
immediately sent to user, and if the video exceeds 100 moving frames, then the
video of those moving frames will be sent in the next e-mail. Motion is detected by
background subtraction method MOG2 algorithm. In the event if the movement is
identified, then human discovery is executed by HAAR cascade classifier. The
proposed framework is sufficiently keen to identify human movement, checks
number of individuals, and informs client by sending SMS, and e-mails the
recorded video to client. Figure 1 demonstrates the block diagram of proposed
framework. Step by step working process is as follows:
(1) Start process by entering correct password. System goes in surveillance mode
as Arduino allows Raspberry pi to turn on the camera and all electrical
appliances are turned off.
(2) Wait for motion detection—Confirm the human detection, people counting
mode and send SMS to the owner, Alarm is turned ON.
774 S.P. Gulve et al.

Fig. 1 Block diagram

(3) Security mode is ON—Record video as security system is broken and e-mail
that recorded video to user.
(4) Enter password again to make system work in normal mode.

3 System Architecture

3.1 Elements of the System

Raspberry Pi2 is the primary handling unit. OpenCV is build and introduced on it
for image processing. Arduino is subprocessing unit which is in charge for ini-
tialization of fundamental handling unit after getting password from client. Python
is utilized for interfacing between Raspberry Pi2 and Arduino, and Python is
additionally utilized for sending SMS through Web; furthermore, it is utilized to
send e-mail to the client/user.

3.2 Hardware and Software Design

In addition to Raspberry Pi, Arduino and GSM module are the principle equipments
utilized for the framework. The GSM module (needs a SIM card to work) is
associated with Arduino by USB to serial converter. A LCD screen is utilized to
show the entered secret word. Two relays of 12 volts are utilized which are asso-
ciated between raspberry pi and Arduino. One relay is in charge of activation and
deactivation of fundamental procedure, i.e., Raspberry Pi. Other relay gives a
control to turn off/on electrical appliances. Figure 2 demonstrates the algorithm
utilized for the proposed framework.
Implementation of IoT-Based Smart Video Surveillance System 775

Fig. 2 Algorithm

4 Results

Figure 3 demonstrates the pictorial view of the proposed framework. For the pro-
posed framework, different components should be dealt with, for example, dis-
tinctive lighting condition intensely aggravates the nature of a camera pictures.
776 S.P. Gulve et al.

Inaccurate decision of parameters selection for various conditions causes issues in

camera vision. Figure 4 demonstrates the password check. On the off chance if the
wrong password is entered, then framework does not get enacted. As framework
peruses right password, it turns off all the electrical appliances present in the pre-
mises. In Fig. 5, red light is a heavy load. Subsequent to perusing right password,
the system turns it off and turns on blue light which indicates the actuation of
primary handling framework, i.e., Raspberry Pi.
After enactment, the framework initiates primary system that identifies move-
ment and confirms human movement. The challenge in image processing of pro-
posed system is to distinguish a human and check number of people.
For this, subtraction of foreground image from the background image is nec-
essary. Figure 6 indicates result for human identification. For execution assessment,
there are few things which ought to be considered, for example, picture handling
time per outline, Web speed, and time required for SMS sending by GSM. Figure 7
indicates result for SMS and e-mail notification.
The recorded video for security mode is in JPEG format as it can be played using
any standard video player. As human motion is detected, main system will count

Fig. 3 Pictorial view of proposed system

Fig. 4 Password verification

Implementation of IoT-Based Smart Video Surveillance System 777

Fig. 5 Controlling electrical appliances

Fig. 6 Human detection

Fig. 7 SMS and e-mail notification

778 S.P. Gulve et al.

number of people and a SMS will be sent to user regarding security alert which will
notify user about unauthorized persons’ presence. GSM module will send a SMS to
user. A video will be recorded and saved on SD card as well as will be e-mailed to
user. Figures 6 and 7 show e-mail and SMS notification results send to user after
human detection. SMS sending is done through GSM and Internet.

4.1 Observations and Comparison

Depending on the experiment performed on the proposed system, Table 1 shows

the observation that is made and Table 2 shows comparison of existing system.

Table 1 Observation table

Parameters Results
Camera speed 30 frames per second
Motion detection speed 3.1–9 s as it checks for 100 moving frames with great
accuracy
Human detection speed 2–5 s
Distance between camera and human 10–15 ft
(for human detection) (depending on .xml file)
Time required to send SMS 20–30 s (depends on Internet speed and Web site
load) for 50 Mbps LAN connection

Table 2 Comparison between existing systems

Reference Hardware and Advantages Disadvantages Application
software
Khot Harish S, IBM Smart Provides Online video Home/office
Gote Swati R, Surveillance front-end streaming surveillance
Khatal Sonali B, Engine (SSE), video which requires
Pandarge camera, Ethernet analysis more Internet
Sangmesh [12] capabilities data usage
U. Ramakrishna, Raspberry Pi, Provides Use of Industries,
N. Swathi [13] USB Web IOT-based low-processing offices/home,
Camera, GSM, smart video power chips military areas,
USB Wi-fi surveillance may result in elderly person
Dongle, HDMI poor falling sick
cable, relay, performance
motion software, speed
Python scripts,
Shell script
Akshada Transmitter and Provides More hardware Office/army/home
Deshmukh, receiver kit, facility of complexity surveillance, bank
Harshalata digital camera multilevel security, space
Wadaskar, Leena security research
Zade, Neha
Dhakate, Preetee
Karmore [14]
Implementation of IoT-Based Smart Video Surveillance System 779

5 Conclusion

The proposed framework is cheaper in cost as it requires less storage space and no
individual to monitor persistently from control room. In the proposed framework,
two hardware boards are utilized to enhance the execution of the framework. The
proposed system also provides facility of instantaneous alert to user so action can be
taken immediately. The proposed system can be implemented at high-alert places
such as banks, industry, or any other places where this type of security is required.
The future thought is to attempt and add some more elements to the framework
like face recognition for user for activation and deactivation of the system and
mobile-based home automation framework which will permit client to control the
proposed framework through mobile. As Internet assumes an imperative part in the
proposed framework, utilization of 3G/4G network would be suggested for better
execution.

References

1. Md. Syadus Sefat, Abdullah Al Mamun Khan, Md. Shahjahan “Implementation of vision
based intelligent home automation and security system”, 3rd INTERNATIONAL
CONFERENCE ON INFORMATICS, ELECTRONICS & VISION 2014.
2. F. Bartolini, V. Cappellini and A. Mecocci, Counting people getting in and out of a bus by
real time image-sequence processing, Image and Vision Computing, vol. 12, no. 1, Jan. 1994,
pp. 36–41.
3. A. Albiol, I. Mora and V. Naranjo, Real-time high density people counter using
morphological tools, IEEETrans. Intelligent Transportation Systems, vol. 2, no. 4, Dec.
2001, pp. 204–218.
4. M. Rossi and A. Bozzoli, Tracking and counting moving people, IEEE International
Conference on Image Processing (ICIP), vol. 3, 1994, pp. 212–216.
5. G. Sexton, X. Zhang and G. Redpath, Advances in automatic counting of pedestrians, 1995
European Convention on Security and Detection (ECSD95), 1995, pp. 106–110.
6. K. Terada, D. Yoshida, S. Oe and J. Yamaguchi, A method of counting the passing people by
using the stereo images, IEEE International Conference on Image Processing (ICIP), vol. 2,
1999, pp. 338–342.
7. O. Masoud and N. P. Papanikolopoulos, A novel method for tracking and counting
pedestrians in real-time using a single camera, IEEE Trans. Vehicular Technology, vol. 50,
no. 5, Sep. 2001, pp. 1267–1278.
8. J. W. Kim, K. S. Choi, B. D. Choi and S. J. Ko, Real-time vision-based people counting
system for security door, International Technical Conference on Circuits/Systems Computers
and Communications, 2002, pp. 1416–1419.
9. J. Bescos, J. M. Menendez and N. Garcia, DCT based segmentation applied to a scalable
zenithal people counter, IEEE International Conference on Image Processing (ICIP), vol. 3,
2003, pp. 1005–1008.
10. T. H. Chen and C. W. Hsu, An automatic bi-Directional passing-people counting method
based on color-image processing, 37th IEEE International Carnahan Conference on Security
Technology, Taiwan, Oct. 2003, pp. 200–207.
11. KaewTraKulPong P, Bowden R. An improved adaptive background mixture model for
real-time tracking with shadow detection Proceedings 2nd European Workshop on Advanced
 780 S.P. Gulve et al.

Video Based Surveillance Systems (AVBS 2001), Kingston, UK, September 2001 3rd
INTERNATIONAL CONFERENCE ON INFORMATICS, ELECTRONICS & VISION.
12. Khot Harish S, Gote Swati R, Khatal Sonali B, Pandarge Sangmesh Smart Video
Surveillance, IJ EERT Volume 3, Issue 1, January 2015, PP 109–112 ISSN 2349-4395
(Print) & ISSN 2349-4409 (Online).
13. U. RAMAKRISHNA, N. SWATHI: Design and Implementation of an IoT Based Smart
Security Surveillance System, IJSETR ISSN 2319–8885 Vol. 05, Issue. 04, February-2016,
Pages: 0697–0702.
14. Akshada Deshmukh, 2 Harshalata Wadaskar, 3 Leena Zade, 4Neha Dhakate, 5 Preetee
Karmore: Webcam Based Intelligent Surveillance System, IJES Vol. 2, 8 March 2013,
Pp 38–42 Issn(e): 2278-4721, Issn(p):2319-6483.

View publication stats