International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072
Smart Traffic System using Machine Learning
UTKARSH BAVISKAR
Department of Information Technology Engineering, VIIT Pune, India.
---------------------------------------------------------***--------------------------------------------------------------
Abstract: - Now days we are in faster time zone. This can The above figure shows the virtual traffic signals which
be growing very faster as per the early days. The contents the four roads and four traffic lights for each. In
technology as well as populations also increasing day by general 30 seconds are allots to every road to clear the
day. According to the largest populations the government traffics. Therefore we are planning to be monitoring the
wants manage or develop the roads, because in the largest each side of traffic by camera and with help of the
population’s country the traffic is biggest issues to be face machine learning. Capturing images or live videos and
by traffic officers. But sometimes single officers can’t processing this by using relevant libraries of python
handle the larger traffic especially in festivals days. which is YOLO. We have different types of vehicles. The
Therefore we are research on those situations. How to ambulance is one of them. We can’t stop for signal to be
manage the traffic automatically? Another situation is clear traffic. Therefore we are needed to learn our
while in larger traffic how to cross the ambulance from project about that difference of vehicles. For training the
that traffic. So this paper is helping us to solve those entire systems the AlexNet CNN architecture are used. The
problems regarding larger traffic. After all the research the steps of working are as follows,
new technology is highlighted i.e. Machine Learning
(ML).We have use some of algorithm, dataset and 1. The camera sending the images to system in some
mathematical calculations which can be done by machine intervals for processing.
learning and python. The python programming language 2. This can be determines the density of traffic from
which can be providing a platform to perform some the roads and based on the calculations time of the
operations like object detections, image processing, and traffic clear is changed which is shown in result.
video processing and so on. We have design some algorithm 3. The system is decided which is signal is open for
which is can handles larger traffic. which time and it will triggered the traffic signals.
Keywords : Smart Traffic, machine Learning, AI, YOLO,
AlexNet.
1. INTRODUCTION
The cities are highly developed with technology and uses
different types of electronics devices, sensors, manages
big data and one of highlighted is city roads and properly
developed. But in these cities the most commonly issue is
facing that can be traffic management. We have seen that
because if some issues largest amount of traffic is created
such as traffic lights not working properly, while
occurring number’s of vehicles arriving on single side of
road as compare with other but the given time is not
sufficient to be clear that side traffic. Therefore such
types of traffic we want to handle very smartly. The
Machine Learning is fulfilling this condition based of
object detection.
2. METHODOLOGY
Fig 2.2:- Flow chart of project
Fig 2.1: Virtual view of Traffic signal
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1979
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072
3. SOFTWARE PLATFORM
● Python
The python is open source programming language it
means you can use this for your own purpose or free to
use anywhere. This is currently in higher demand to IT
industry. It is mostly use for the Machine learning to
develop the websites, data science and software’s etc. The
python is very friendly to everyone. The python is almost
similar to the C language only the difference in coding
syntax. It is capable to perform various types of
operations and it use to build the machine learning, data
Fig. 4.2: Layer Views
analyze, complex statistical calculations.
The python has number‘s of libraries support to perform
This is the python library which is called “You Only Look
varieties of task. Therefore we are choosing this language
Once”. It is “real time object detections recognition
for developing this project. If you want to use any kind of
system”. The YOLO is a single neural network.
libraries you need to be installing in your desktop. “pip
install ” is the command to install any libraries of python.
5. Alex Net’s ARCHITECTURE
● Machine learning
The Machine learning (ML) is nothing but the artificial
intelligence (AI). This is software application. It is has
ability to automatically learn and improve the
performance without making a complex programming.
“Machine learning focuses on the event of pc programs
which will access information and use it to find out for
themselves”.
“The ML permits analysis of huge quantities of data.
Whereas it usually delivers faster, additional correct
leads to order to spot profitable opportunities or
dangerous risks, it should additionally need extra time Fig. 5.1: AlexNet Architecture Diagram
and resources to coach it properly. Combining machine
learning with AI and psychological feature technologies We are using AlexNet CNN Model architecture to
will build it even simpler in process large volumes of streaming videos. Basically we are using this for
information” [1]. detecting the various types of Ambulance from video. For
performing these operations we are training to or system
4. YOLO to determine the Ambulance with help of dataset. In our
dataset we are storing different types of vehicles image.
This can help to our systems all detect all the cars
motorcycles and so on.
● Convolution Layer (96, 11×11)
● Max Pooling Layer (3×3)
● Convolution Layer (256, 5×5)
● Max Pooling Layer (3×3)
● Convolution Layer (384, 3×3)
● Convolution Layer (384, 3×3)
● Convolution Layer (256, 3×3)
● Max Pooling Layer (3×3)
● Fully-Connected Layer (4096)
● Fully-Connected Layer (4096)
Fig. 4.1: Object detections ● Fully-Connected Layer (1000)
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1980
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072
8. RESULT
6. CONVOLUTION NEURAL NETWORK
Fig. 6.1: Convolution Operation Example
7. DATASET
Fig 8.1:- Smart Traffic Control system
Basically the Dataset which we have used it is an
manually created dataset in this dataset we have
collected the Emergency Vehicle images and Non
Emergency Vehicle Dataset and we have gathered the
images from Google, So in this dataset we have collected
1380 Emergency Vehicle Images and 1,496 Non
Emergency Vehicle and further while training we have
divide this dataset into training and testing part.
Fig 8.2:- Camera View of all side
Fig. 7.1 : - Dataset 1
9. ADVANTAGES
● The uses of this system make our cities as smartest
cities.
● It has capabilities to handle traffic as well as monitor
the traffic.
Fig. 7.2 : - Dataset 2 ● The main advantage of this system is helping the
ambulance vehicle. This can be happens by clearing
the traffic from ambulance side roads.
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1981
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072
10. CONCLUSION
In this way we are developing a very smart traffic control
system which can be able to detect and monitor the
traffic. It can be take decision according to the density of
traffic.
11. REFERENCES
[1] M.Wieringet,"Multi-agent reinforcement learning for
traffic light control", inICML, pp. 1151-1158, 2000.
[2] Y. K. Chin, N. Bolong, A. Kiring, S. S. Yang and K. T. K.
Teo, "Q-learning based traffic optimization in
management of signal timing plan", International
Journal of Simulation Systems Science & Technology,
vol. 12, no. 3, pp. 29-35, 2011.
[3] S. El-Tantawy and B. Abdulhai, "Multi-Agent
Reinforcement Learning for Integrated Network of
Adaptive Traffic Signal Controllers (MARLIN-
ATSC)", 2012 15th International IEEE Conference on
Intelligent Transportation Systems, Sep. 2013.
[4] Q Wang, J Wan, Y Yuan, “Locality constraint distance
metric learning for traffic congestion” Pitu, Larry
Head, A real-time traffic signal control system:
architecture, algorithms, and analysis, NOV 2016
[5] X. Liang, X. Du, G. Wang and Z. Han, "Deep
Reinforcement Learning for Traffic Light Control in
Vehicular Networks", arXiv:1803.11115, vol. 1, Mar.
2018.
[6] H. Wei, G. Zheng, H. Yao and Z. Li, "IntelliLight",
Proceedings of the 24th ACM SIGKDD International
Conference on Knowledge Discovery & Data Mining -
KDD 18, 2018.
[7] MayuriLakhani, Shivansh Dave, Jignesh J.Patoliya
“Internet of Things based Adaptive Traffic
Management System as a part of Intelligent
Transportation System (ITS)”, IEEE 2018.
[8] AditiYadav, Vaishali More, NehaShinde,
ManjiriNerurkar, NitinSakhare, “Adaptive Traffic
Management System Using IoT and Machine
learning”, Journal of Electrical Technology, Vol 4,
pp.221-226, Jan 2019.
[9] Dr.A.Ravi,R.Nandhini, Bhuvaneshwari, J.Divya,
K.Janani,“ Traffic Management System using Machine
Learning Algorithm” , INTERNATIONAL JOURNAL OF
INNOVATIVE RESEARCH IN TECHNOLOGY, Volume 7
Issue 11, April 2021.
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1982
