An IoT Concept for Region Based Human

Detection Using PIR Sensors and FRED Cloud

Naba Krushna Sabat Umesh Chandra Pati
Dept. of Electronics & Communication Engineering Dept. of Electronics & Communication Engineering
National Institute of Technology National Institute of Technology
Rourkela, India Rourkela, India
[email protected] [email protected]

Biswa Ranjan Senapati Santos Kumar Das

Dept. of Computer Science & Engineering Dept. of Electronics & Communication Engineering
National Institute of Technology National Institute of Technology
Rourkela, India Rourkela, India
[email protected] [email protected]

Abstract—Present day Wireless Sensor Networks small in size low-cost and uses low-power hence it can
(WSNs) and Internet of Thing (IoT) plays a vital role for deploy easily anywhere. This is used both in indoor as
Surveillance and security. This paper presents region based well as outdoor application for detecting the object.
object/human detection and alert notification to the control In this paper, we modeled a technique using PIR
room or administration. In this case PIR sensors are used in
sensor and NodeMCU to detect object/human in region
different region which helps to detect and alert the presence
of objects covered by the sensor zone which is called as the wise and counts number of objects passes through the
field of view (FoV). An ESP-8266 Wi- Fi module is used to region. Here it also shows how the information is sent to
create the WSN network. It helps to send the information to FRED cloud service and an audio device is used gives the
the cloud using cloud service provider. The information is alert siren that object detected in the particular region.
sent to the FRED cloud service and the same can be stored in The rest of this paper is organized as follows. Section
local database (MongoDB). This information can be II illustrate the literature survey related to PIR sensor for
retrieved from the database at the time of requirement i.e. object detection. Section III describes the System model.
for future use. From the implementation result, it has been Section IV presents the proposed method. In Section V
found that with proper deployment of sensors, PIR sensors
the implementation and discussion is explained in details
perform well for the human/object detection.
with the flow chart and algorithm. Finally, we present the
Keywords—PIR sensor, ESP8266 Wi-Fi module, Cloud, conclusion in Section VI.
MongoDB, Node-Red, FRED, IOT.
II. LITERATURE SURVEY
I. INTRODUCTION In [7] they proposed a method of Intrusion detection
Security and surveillance is an important factor using PIR sensor with IoT concept. They use sensors,
playing in present scenario. For that different types of controllers, GSM modulated and Thing Speak cloud
techniques are proposed. PIR sensor is a motion detective service to detect and alert of intruder. PIR sensor nodes
sensor also used for detecting and alerting the presence of are mounted on ceiling for human indoor localization [8].
objects in a particular region covered by the sensor zone This paper describes how the human target detect
called field of view (FoV). In other words, it can be say independently comparing to other existing technique.
that providing security and surveillance of that area. They use kalman filter kalman smoother for human
Internet of Thing (IoT) also have an influence factor in estimation of human position [8], whereas in [2] used
day-to-day life, it is a fast growing technology where all machine learning technique (i.e. feature extraction and
things are connected to a network and communication sensor fusion) with PIR array sensors for detecting,
done between things, human to human, things to human tracking people direction in hallway also measures the
and vice-versa [1]. The main moto of IoT enabled thing is distance with an accuracy of 83.49% - 95%.
that the sensors are connected using different technology Some paper uses different algorithm and implement
may be wired or wireless, and information of the sensor with PIR sensor for human tracking. B. Song et.al.
can send anywhere at any time [2]. For example, by using proposed region based tracking algorithm for human
PIR and NodeMCU, detection of vehicles and tracking [9]. Q. Hao et.al. designed a sensor network
transmission of critical environmental parameters to the using PIR and Fresnel lens array for distributed multiple
vehicle can be possible in VANET [3]. human tacking [6]. They used joint probabilistic
PIR sensor is working based on the principle of association and consensus filtering method for multiple
thermal detectors. Any absorbing material comes in the human tacking. Lai et al. proposed two approaches for the
FoV of PIR causes change in temperature, then the detection of location of residents such as terminal based
detector analyze both of object. The output depends on and non-terminal based methods [4].
different characteristic like sensing range, object moving,
distance of object FoV of sensor etc. [5]. PIR uses Fresnel
lens which helps for designing FoV, also have high
sensitivity for a distance of 2-10 meter [6]. This sensor is

978-1-7281-0419-5/19/$31.00
Authorized ©2019
licensed use limited to: Henri IEEEAir Force Academy Brasov. Downloaded on February 05,2024 at 13:51:31 UTC from IEEE Xplore. Restrictions apply.
Coanda
 III. SYSTEM MODEL coverage area is 1100 in horizontal whereas 930 in
This section discussed about different hardware and vertical. The Fig. 3 shows the detection range of
software Horizontal (TOP view) and Vertical (Side view)
 Node-red: Node-red is a visual tool developed by according to the data sheet [9].
IBM Technology and is used for wiring the IOT.
Here the inputs, outputs, and controllers are
connected and provide flow to process data, connect
things and sends information to the cloud or sends an
alert message as required. Node-red used for
designing front-end with the help FRED (Front end
for node-red) tool, FRED manages instances for node
red for multiple users in the cloud.
 NodeMCU: NodeMCU ESP8266 is a 30-pin
microcontroller have a self-containing Wi-Fi module.
From the datasheet, it found that it has 17 number of
GPIO (General Purpose Input and Output) pins,
where the PIRs are connected. Here ESP8266 Wi-Fi
module is programmed with Arduino IDE for
detecting Object in different Region. Figure 1 shows Fig. 3 Top and bottom view of PIR
the pin diagram of NodeMCU [10].
Fig. 4 shows the complete block diagram of region
wise object detection. Here the complete region is divided
into four sub-regions according to the coverage area range
of PIR sensor. NodeMCU act as the heart of the module.
Object detected in each region is sensed by the respective
PIR sensor and information send to cloud as well as the
database with the help of Wi-Fi module.

IV. PROPOSED METHOD

The proposed model describes how a PIR sensor
array can detect objects region wise and can send or alert
information to the Authorization center. Here it can say
Fig.1 Pin diagram of node-MCU that using the PIR array module one can cover all the
region and can detect any type of object/Human in the
 Passive Infra-Red (PIR): Passive infrared radiation particular area and is alert as well as send information to
(PIR) sensor is an electronic sensor which detects the cloud.
objects from the infrared radiation. Each object
generates some heat which interns to infrared
radiation. So when the object has moved in the
coverage area of the sensor there is a change in
infrared radiation causes detection of objects (like
Human, animal, car, etc.). The sensing time can be
adjusted as per the requirement there are two
adjustable knob is present, one to increase or decrease
the sensitivity and another for time delay as shown in
Fig. 4 Block diagram of region wise object detection
Fig. 2 [7].

Fig. 2 Pin diagram of PIR

PIR also have some limitation like detection range

and coverage area. From the datasheet, it says the
maximum distance of the PIR sensor is 10m, and the Fig. 5 Flowchart

Authorized licensed use limited to: Henri Coanda Air Force Academy Brasov. Downloaded on February 05,2024 at 13:51:31 UTC from IEEE Xplore. Restrictions apply.
 In this proposed model two PIR sensor is covered two object detects the LED indicates immediately and retain in
regions and detect the human passes on the field of the the same state for at least one to two second. This causes
region. Once an object comes into the region there is a due to response came from the PIR. One can say this is
change in infrared radiation when there is a change in the one of the limitations in PIR sensor that even if the Object
radiation of signal detected by the sensor; it detects there passes over the ROC (Region of Coverage) area, the PIR
is an object and the output pin of PIR goes high which detects there is a presence of object because of delaying
intern glows the LED connected to particular PIR of that time response signal. Fig. 6 shows some of the practical
region. Fig 5 shows the flow chart design of the proposed result test in our laboratory. In this section we have some
model. implementation results of node-red and FRED connection.

Algorithm
The pseudo code of the proposed algorithm is
mentioned in Algorithm 1.
Algorithm 1 Region-wise Human Detection using
PIR Sensors

1: Input: Presence of human/object in any region.

2: Output: Detection and count the number of human in
each region.
3: Sense the presence of human using PIR sensor
module.
4: if Human is detected then
5: Identify the detection region
6: if Detection is in Region-1 then
7: LED-1 corresponding to region-1 will glow.
8: Increment the counter for the corresponding
region-1 detection.
9: else
10: No change in the state.
11: end if Fig. 6 Implementation result
12: if Detection is in Region-2 then
13: LED-2 corresponding to region-2 will glow. Fig. 7 shows how the connection made by the help of
14: Increment the counter for the corresponding node-red nodes to FRED cloud service. Different nodes
region-2 detection. used in the Fig. 7 are timestamp which initiates the flow
15: else with a specific payload, http request - to send the request
16: No change in the state. and to return its response from a server i.e. NodeMCU,
17: end if switch-to forward the message based on matching defined
18: Send information to the cloud. rules, function - to send the message to the FRED cloud
19: else service, counter- to count the number of human in a
20: Wait for the detection of the human. region, and FRED cloud node is used to send the
21: end if information to the FRED cloud service. Fig. 8 shows
22: STOP

V. IMPLEMENTATION AND DISCUSSION

PIR detect the object and sends information to
ESP826 Wi-Fi module. The NodeMCU has programmed
in such a way that it sends information to the cloud with
the help of node-red platform as shown in figure. FRED is
used here as a cloud where it shows the number object
passes through a particular region, i.e. In Region-1 and
Region-2. To publish the information, here the NodeMCU
used as a server, and the NodeMCU generates its IP. This
Fig. 7 Node red wiring connection for detection
IP is used to publish the information in FRED cloud.
FRED provides a wide range platform where data can the FRED frond end connection. In Fig. 9 the dash board
store and publish both in public as well as private. The shows the details of object like number of object counts,
information is also stored locally in the MongoDB notification message etc.
database. Here we make it public so that all can access the
information. Once it is privatized only the authenticated
person can access the cloud service.
In practical test of our laboratory, it was found that
PIR can cover the range 900 when used in Horizontal with
a distance of 7.3 meters. It has been seen that once the

Authorized licensed use limited to: Henri Coanda Air Force Academy Brasov. Downloaded on February 05,2024 at 13:51:31 UTC from IEEE Xplore. Restrictions apply.
 delivered and takes necessary steps for further
requirement.
ACKNOWLEDGMENT
This research was supported by the project
“Development of Wireless Multi Sensor based Intelligent
Security & Surveillance System” funded by DRDO.

REFERENCES
[1] O. Vermesan, P. Friess et al., Internet of things-from research
and innovation to market deployment. River publishers Aalborg,
2014, vol. 29.
Fig. 8 Snap shot of Fred cloud connection for object detection [2] M. S. Mahmoud and A. A. Mohamad, “A study of efficient
power consumption wireless communication techniques
techniques/modules for internet of things (IoT) applications,”
2016.
[3] B. R. Senapati, R. R. Swain, and P. M. Khilar, “Environmental
monitoring under uncertainty using smart vehicular ad hoc
network,” 2018.
[4] K.-C. Lai, B.-H. Ku, and C.-Y. Wen, “Using cooperative pir
sensing for human indoor localization,” in 2018 27th Wireless
and Optical Communication Conference (WOCC). IEEE, 2018,
pp. 1–5.
[5] P. Zappi, E. Farella, and L. Benini, “Tracking motion direction
and distance with pyroelectric ir sensors,” IEEE Sensors
Journal, 2010, pp. 1486–1494.
[6] Q. Hao, F. Hu, and J. Lu, “Distributed multiple human tracking
with wireless binary pyroelectric infrared (PIR) sensor
Fig. 9 Dashboard for object detection in region-wise.
networks,” in SENSORS, 2010 IEEE. IEEE, 2010, pp. 946–950.
[7] K. C. Sahoo and U. C. Pati, “IoT based intrusion detection
VI. CONCLUSION system using pir sensor,” in 2017 2nd IEEE International
This paper describes the design and implementation Conference on Recent Trends in Electronics, Information &
of IOT based object/human detection in region-wise using Communication Technology (RTEICT). IEEE, 2017, pp. 1641–
PIR sensors. ESP8266 Node MCU used as the heart of the 1645.
[8] X. Luo, T. Liu, B. Shen, L. Gao, X. Luo et al., “Human indoor
circuit which leads to send information to the cloud. Once localization based on ceiling mounted pir sensor nodes,” in 2016
motion encounters in the coverage area of the sensors, a 13th IEEE Annual Consumer Communications & Networking
POP-UP message shows in the FRED cloud and also Conference (CCNC). IEEE, 2016, pp. 868–874.
detects the number of objects in the particular region. This [9] B. Song, H. Choi, and H. S. Lee, “Surveillance tracking system
using passive infrared motion sensors in wireless sensor
can be helpful for intrusion detection by using siren also network,” in 2008 International Conference on Information
sent an alert mail to the administrator. This circuit is Networking. IEEE, 2008, pp.1–5.
cheap to design, easy to install and consume less power. [10] https://www.theengineeringprojects.com/2018/10/introduction-
Finally, users can alert once the message has been tonodemcu- v3.html.

Authorized licensed use limited to: Henri Coanda Air Force Academy Brasov. Downloaded on February 05,2024 at 13:51:31 UTC from IEEE Xplore. Restrictions apply.