Proceeding of 2018 IEEE International Conference on Current Trends toward Converging Technologies, Coimbatore, India

Density Based Smart Traffic System with Real

Time Data Analysis Using IoT
Naga Harsha.J Nikhil Nair
School of Electrical Sciences School of Electrical Sciences
Karunya Institute of Technology and Sciences, Karunya Institute of Technology and Sciences,
Coimbatore, India Coimbatore, India
[email protected] [email protected]

Sheena Mariam Jacob J.John Paul

School of Electrical Sciences School of Electrical Sciences
Karunya Institute of Technology and Sciences, Karunya Institute of Technology and Sciences,
Coimbatore, India Coimbatore, India
[email protected] [email protected]

Abstract—This paper aims to develop a convenient unwanted congestion in traffic. The major difference in
traffic system that allows a smooth movement of cars most of the previous models lies in the type of system and
which will help build a smarter city. The traffic system sensors used to calculate the density of traffic in a specific
currently implemented in many areas is not based on lane. They all aim to overcome the disadvantage of the
the density of traffic and every road is allotted a preset traditional system, and use various sensors and algorithms
time. This results in traffic congestion due to large to implement a smarter system. The different sensors that
red-light delays and timings allotted for roads in a city are widely used to find the density of traffic include
that should vary during peak on-off hours, but in Inductive loop detectors, Piezoelectric sensors[1],
reality don’t. These traditional systems are not Ultrasound sensors [2], Infrared sensors, Sound
adaptable and fail to support traffic during an sensors[3], Acoustic Sensors[4],Image processing
unexpected situation or during an accident, and this algorithms[5] based on live feed from a camera, and many
makes them inefficient. In order to calculate the others. Our proposed system aims to calculate the density
density of traffic various sensors can be used, each of traffic by using a combination of ultrasound sensors
having their merits and demerits. In our proposed and a camera to find the level of traffic in a lane. These
system Ultrasound Sensors are used along with Image levels are classified as low, medium ,and high and is
Processing(using live feed from a camera) that works dependent on the length of the road and the density of
on a Raspberry Pi platform and calculates the vehicle traffic in that area. The values indicated by the sensor and
density and dynamically allots time for different levels the processed density value of the image are in turn sent
of traffic. This in turn allows better signal control and to a Raspberry Pi which sends the values to the cloud. The
effective management of traffic thereby reducing the Internet of Things is used to communicate with the cloud
probability of a collision. By using Internet Of for the real time analysis of data.
Things(IoT) real time data from the system can be
collected, stored and managed on a cloud. This data
can be used to interpret the signal duration in-case any
of the sensing equipment fail, and also for future
analysis.

Keywords: Density based system, Raspberry Pi,

Ultrasound sensors, Image Processing, Internet of
Things(IoT), Real time analysis, Cloud.

I.INTRODUCTION
Fig 1 Smart Traffic Light
Urban mobility has become one of the key sectors of
smart city development that has rapidly spread across This data will be useful when the sensor system fails and
India. Over the last few years there have been umpteen will serve as a backup system while also providing
papers presented based on the implementation of a smart information for future analysis of traffic. By allotting time
traffic management system to counter the traditional based on the levels of traffic vehicles are prevented from
preset time period system, which causes most of the waiting for long periods thereby reducing the pollution

978-1-5386-3702-9/18/$31.00 © 2018 IEEE 1

zed licensed use limited to: SVKM¿s NMIMS Mukesh Patel School of Technology Management & Engineering. Downloaded on August 20,2021 at 06:40:57 UTC from IEEE Xplore. Restrictions
 Proceeding of 2018 IEEE International Conference on Current Trends toward Converging Technologies, Coimbatore, India

that is caused. The system is cost effective, saves time, on the sides of the roads and also on the road, below the
and can also prevents collisions. Unlike other systems, it vehicles. Unlike other sensors ultrasound sensors are not
also has a fail safe system that will be used in case there affected by rainfall as ultrasound travels in water. It also
are errors in the sensor reading’s. Fig 1 shows a smart does not require approval from the FCC. The HC-SR04
traffic light system ultrasonic sensor, as shown in Fig.3 is used to detect
vehicles.
II.SYSTEMS IN EXISTENCE

A. The Traditional System

The traditional traffic system was developed in the 20 th
century, and the time for which the signal was set to run
was decided beforehand. Every lane had a predetermined
time and irrespective if there was traffic on a lane the
traffic indicator would still count down from its preset
value, resulting in heavy traffic on other lanes. This
system is still followed in a number of countries, and is
inefficient and results in poor time management.

B. Density Based Traffic System

Fig 3. Ultrasonic Sensor HC-SR04
The density based traffic system calculates the amount of
traffic on a specific lane using vehicle detectors or
3) Camera Module :
sensors. In countries where there exists bumper to bumper
Image processing is one of the most efficient ways of
traffic, designing a smart traffic management system is
measuring traffic density in real time, and have already
highly important so as to prevent accidents and save time.
available algorithms as depicted in [3],[6] that can be used
The efficiency of the systems is dependent on how
for monitoring vehicle movement in a lane. Traffic can be
efficient the sensors are. The various sensors used for
monitored using still images[7] or images from a video.
vehicle detection include:
Due to the fact that the cameras are mounted above the
road, the installation process is easy and convenient. Fig 4
1) Infrared sensors:
shows a camera module that can be used with the
Infrared sensors whether active or passive can be used to
Raspberry Pi.
find a vehicle’s count, presence, and occupancy. An
active infrared sensor can also be used to find the speed of
a vehicle. While they have the advantage of being able to
respond throughout the day and night, they are affected by
light, smoke, dust and fog and have a very short range.
Fig 2 shows an IR Sensor that has a IR emitter and
receiver.

Fig 4. Camera module with the raspberry pi used for

image processing

4) Inductive Loop Detector:

Inductive Loop Detectors consists of a number of loops
Fig. 2. IR Sensors which are integrated in a pavement and connected to a
control box. When a car passes over it creates an EMF
2) Ultrasound sensors : (electro magnetic field) thereby reducing the inductance
Ultrasound sensors are used to determine the presence of of a loop and this is recorded by a meter as shown in Fig
an obstacle and are the most widely used sensors in Japan 5. These loops are used for vehicle detection and count,
and a number of other countries. They can be placed on and has been implemented in about 90% of all the traffic
the road, and do not have to be embedded as in the case of signals in the USA. They are the most consistently
piezoelectric and inductive loops. The positioning of the accurate detectors but sources of loop malfunction can
ultrasound sensors also makes a difference to the produce erroneous data, and may lead to inaccurate
detection of traffic [2]. Ultrasound sensors can be placed detection. These systems have more complexities, and

zed licensed use limited to: SVKM¿s NMIMS Mukesh Patel School of Technology Management & Engineering. Downloaded on August 20,2021 at 06:40:57 UTC from IEEE Xplore. Restrictions
 Proceeding of 2018 IEEE International Conference on Current Trends toward Converging Technologies, Coimbatore, India

have to embedded in the ground. 5) Piezoelectric sensors:

The piezoelectric sensor detects the motion of the car due
to the pressure applied when a car passes over the sensor.
This means that the car’s wheel has to be aligned above
the sensor in order for it to detect vehicles at a particular
position. This makes it useful for only dynamic
measurement.

A comparison of the various sensors with its advantages

and disadvantage are given in Table I.

Fig 5. Working of inductive loops for vehicle detection

TABLE I.COMPARISON BETWEEN VARIOUS SENSORS USED FOR MONITORING TRAFFIC [1]
Type Merits Demerits
Passive Acoustic 1. Provides 24 hours operation 1. Rather new technology and hasn’t been
Detector array 2. Not affected by weather conditions. investigated thoroughly.
3. Passive monitoring technique so no
harmful signals are transmitted.

Piezoelectric Sensor 1. Can be used to find the number and type of 1. For permanent installation must be
vehicles on a traffic lane. embedded in the road, and would
2 .Also detects when and where a vehicle therefore require replacement due to
passes by. potholes, and other issues.
3 .They can be also used to determine the 2. The wheel of the vehicle should be right
speed of the vehicle above the sensor in order for detection to
take place.

Ultrasound Sensor 1. Effective in finding the level of traffic in a 1. It is sensitive to change in temperature
certain lane 2.Reflection from thin, curved surfaces
2.All weather operation and it is not affected will not be accurate.
by rainfall, dust or snow.
3.It can sense all material types
4.Does not need FCC approval

Inductive Loop 1.When properly installed, provide the most 1.They cannot be used to directly measure
Detector accurate data for detection in all weather the speed of the vehicle.
condition. 2.Installation will result in traffic
2.The design of loops is flexible and it can be disruption
used for a wide range of applications. 3. While repairing roads, it may be
required to re- install the loop.

Microwave/Millimeter 1. Single detector can cover multiple lanes. 1. False detection due to multipath
wave radar 2. FCC approval not required if it is to work propagation.
in the Ku or X band.

Magnetic Detector 1. In situation where inductive loops cannot 1. Magnetic Detectors will have to be
(Magnetometer) be implemented, magnetometers serve as an embedded in the pavement.
efficient replacement. 2.Lane closure is required in order set the
2.They are small self contained units and can ambient magnetic conditions
therefore installation requires less ground 3.Can only be used for small area’s.
work than loops. 4. Multiple detectors are required for
3. Long lasting battery life is another major smaller vehicle detection.
attribute.

zed licensed use limited to: SVKM¿s NMIMS Mukesh Patel School of Technology Management & Engineering. Downloaded on August 20,2021 at 06:40:57 UTC from IEEE Xplore. Restrictions
 Proceeding of 2018 IEEE International Conference on Current Trends toward Converging Technologies, Coimbatore, India

Video/image 1.Mounted above the road, so installation 1. Affected by shadows, weather, and
Processing process is easy reflections from the roadway surface.
2.Image processing algorithms for vehicle
detection are already available.
3. Provides wide-area detection.
Infra Red Sensor 1.Infrared Sensors have a longer lifetime due 1.Infrared frequencies are affected by
to the fact that they consume less power. smoke, dust, fog, sunlight, etc.
2.Reliable detection of motion during day, as 2.The efficiency of the sensor degrades
well as night. with longer distances
3.Accuracy is not affected by oxidation or 3.For a surface, it is required to know the
corrosion. nature of reflection, scattering and
absorption to measure the distance
effectively.[8]

C. Smart Traffic System Based on IoT: computer which in turn is connected to the cloud.
Due to the fact that IoT is a relatively new concept there Information from these sensors is collected and analyzed
is no exact definition for the Internet of Things, but in by the computer, there by assigning appropriate signal
general terms it can be defined using this equation [9] : duration’s.

IoT = physical entity + controllers, sensors, actuators

+ internet.
C. Flowchart

The advance in technology, and the increase in the

number of sensors in almost every application, has
resulted in large amounts of data that is to be collected,
processed, stored and presented in an easily interpretable
form. Using IoT data that would take months to process,
is just a click away. [3],[10] portray different ways IoT
can be used in to achieve its goal.

III. PROPOSED SYSTEM:

A.Block Diagram

B. Working
Three ultra sound sensors [Fig 3] are placed at fixed
locations having equal distances(calculated based on the
length of the road, and the traffic density in the region) on
the road. The first sensor is for low, second for medium
and third for high density of traffic. A camera is used to
view the traffic on the road, and the images collected from
it are processed using suitable algorithms. These sensing
equipment are interfaced with a Raspberry pi [Fig 4]

zed licensed use limited to: SVKM¿s NMIMS Mukesh Patel School of Technology Management & Engineering. Downloaded on August 20,2021 at 06:40:57 UTC from IEEE Xplore. Restrictions
 Proceeding of 2018 IEEE International Conference on Current Trends toward Converging Technologies, Coimbatore, India

attention. In such scenario’s the computer stops collecting

data form the sensors and accesses the previously stored
values from the cloud.

IV. CONCLUSION

By using this system configuration the possibilities of

traffic jams caused by the currently implemented traffic
system is reduced. This will in turn decrease the workload
of officers, who have to direct traffic in case the traffic
lights are not responding ,or in the case of an event. The
system has a low probability of malfunctioning and even
if it does, a fail-safe system can be initiated by getting the
average density in that area for a particular time frame
from the cloud. Continuous monitoring of traffic can also
be done simultaneously

V.FUTURE SCOPE

The project can further be extended to transfer the values

from the cloud to an application, so that the data collected
by the system is accessible to the users. Be it people who
just want to check the amount of traffic on a road, or for
those who want to study the change in traffic densities
over the years this system can prove to be highly
beneficial to everyone and not just to the traffic
authorities.

REFERENCES

[1]. Piotr Burnos, Janusz Gajda, Piotr Piwowar, Ryszard

Sroka, Marek Stencel, and Tadeusz Zeglen,
“Measurements of Road Traffic Parameters Using
Inductive Loops and Piezoelectric Sensors,” Metrology
and Measurement Systems, vol. 14, no. 2, pp. 187–203,
2007.
[2]. Ashish Jain, Manisha Mittal, Harish Verma, and Amrita
Rai, “Traffic Density Measurement based On-road
Traffic Control using Ultrasonic Sensors and GSM
Technology,” in Proc. of International Conference on
Emerging Trends in Engineering and Technology.
[3]. G. Lakshminarasimhan, V. Parthipan, Mohammed Irfan
Ahmed, Sri Harsha K Nvm , Dr. D. Dhanasekaran,
“Traffic Density Detection and Signal Automation using
IoT,” International Journal of Pure and Applied
Mathematics,Volume 116 No. 21 2017, 389-394.
[4]. Yueyue Na, Yanmeng Guo, Qiang Fu, and Yonghong
Yan, “An Acoustic Traffic Monitoring System: Design
and Implementation,” UIC-ATC-ScalCom-CBDCom -
For example consider our system has the three sensors IoP.2015.41
placed 50 meters apart. In the case of low density the [5]. Sk Riyazhussain, Riyazhussain, C. R. S. Lokesh, P.
allotted duration is 20 seconds, for medium density 40 Vamsikrishna, Goli Rohan, “Raspberry Pi Controlled
seconds and for high density 60 seconds. The data Traffic Density Monitoring System,” IEEE WiSPNET
2016 conference.
received by the computer is stored on a cloud. The
[6]. Payal Gupta, Dhananjay V. Gadre, Tarun Kumar Rawat,
information from the ultrasound sensors is compared with “Real Time Traffic management Control System,” in the
that of the camera continuously. The time duration is International Journal of Electronic and Electrical
allotted to the signals only if both the data match. If not Engineering,Volume 7, Number 5 (2014).
the computer notifies the maintenance team that there is [7]. Dinkar Sitaram, Nirupama Padmanabha, Supriya S,
an error in the sensing equipment and needs immediate Shibani S, “Still Image Processing Techniques for

zed licensed use limited to: SVKM¿s NMIMS Mukesh Patel School of Technology Management & Engineering. Downloaded on August 20,2021 at 06:40:57 UTC from IEEE Xplore. Restrictions
 Proceeding of 2018 IEEE International Conference on Current Trends toward Converging Technologies, Coimbatore, India

Intelligent Traffic Monitoring,” 2015 Third International

Conference on Image Information Processing
[8]. P. M. Novotny, N. J. Ferrier, “Using infrared sensor and
the Phong illumination model to measure distances,”
International Conference on Robotics and Automation,
Detroit, MI, vol. 2, April 1999, pp. 1644- 1649.
[9]. Mahesh Lakshminarasimhan, “Advanced Traffic
Management System Using Internet of Things”, March
2016
[10]. Tanvi Tushar Thakur, Ameya Naik, Sheetal Vatari,
Manjiri Gogate, “Real Time Traffic Management using
Internet of Things,” International Conference on
Communication and Signal Processing, April 6-8, 2016,
India

zed licensed use limited to: SVKM¿s NMIMS Mukesh Patel School of Technology Management & Engineering. Downloaded on August 20,2021 at 06:40:57 UTC from IEEE Xplore. Restrictions