Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

International Journal of Advance Science and Technology

Vol. 29, No. 10S, (2020), pp.3698-3706

Traffic Violation Detection Using Image Processing


C.Nikhil1,Charan Murthy MS 2,Chandan reddy B 3 ,Dinesh Chandra 4and
Vidyasagar KN 5
1
Department of ECE REVA University Bangalore, India,
Department of ECE REVA University Bangalore, India
3
Department of ECE REVA University Bangalore, India
4
Department of ECE REVA University Bangalore, India
Department of ECE REVA University Bangalore, India
1
nikhilcnr06@gmail.com,2chiru.ms1998@gmil.com3R16EC058@ece.reva.edu.in,
4
R14EC044@ece.reva.edu.in, 5vidyasagarkn@reva.edu.in
Abstract

Road safety is often neglected by riders worldwide leading to accidents and deaths.
Usually, road accidents mainly take place due to traffic violation and people lose their
lives in these accidents. The policeman is facing various problems regarding the
maintenance of the database of various violated vehicles number and the violated rules.
The main aim of this paper is to observe the traffic and detect traffic violations such as
helmet detection, signal jump detection and also detects the number plate more quickly
when the violation is detected it records the vehicles licensed number plate and sends the
message to the registered number with vehicles licensed number plate. This article
discusses the various systems built for traffic violations and detection, system overview,
and system design. It also discusses the application of the system.

Keywords: Traffic violation, helmet detection, GSM Module, MatLab, signal jump
detection, Camera, image processing.

1. INTRODUCTION

Technologies play a vital role in making life simpler. In the entire nation, approximately
two lakh people lose their lives due to road accidents every year. By implementing traffic
violations and detection technology helps to detect the violation automatically without the
requirement of human power. This system detects if a person is not wearing a helmet and
also a signal jump which are the main reasons for road accidents. After detecting the
violation it records the number plate and sends the alert message to the vehicle owner
which gives caution to the rider that traffic violation has been made and particular traffic
violation fines have been penalized on vehicles licensed number plate.

2.LITERATURE SURVEY

R Shreyas [1] in their article discusses how a traffic violation can be detected by
identifying the number plate of the vehicle and send them the message for the identified
owner based on the GSM Module. Dharma Raj KC [2] in their article discusses how the
helmet violation can be detected using image processing and deep convolutional neural
network techniques. Kunal Dahiya [3] in their article discusses how a system can assist
the traffic police by detecting the bike rider who is not wearing a helmet using the video

ISSN: 2005-4238 IJAST 3698


Copyright ⓒ 2020 SERSC
International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp.3698-3706

surveillance technique in real-time. XiaolingWang[4] in their article discusses how traffic


violations can be detected in the moving vehicles by feature-based tracking methods
which gives the accurate results of detection and tracking of the traffic violations.
XiaopengJi[5] in their article discusses the methods of target detection and recognition
and a standard of how to set traffic congestion according to traffic conditions. The results
show effectiveness in reducing the false detection rate and improving the detection speed.
RhenAnjeromeBedruz[6] in their article discusses the model used two robots, one for the
moving object which will be detected by the camera, and one for the robot that will follow
the moving object if its speed reaches a certain limit. Siddharth Tripathi[7] in their article
an intelligent system called CBITS. The results discuss mainly three main functions:
emission monitoring, accident detection, and unique vehicle identification.

Seda Kul[8] in their article discuss how to control the traffic flow, vehicle detection,
tracking, security lane, and wrong way violation detection. Chaithra B [11] in their article
discusses Automatic Number Plate

Recognition system designed to detect the vehicles that violate traffic signals by
extracting their number plate from digital images and it sends a violation SMS to the
owner of the vehicle immediately.

3.METHODOLOGY

Traffic Violation and detection consists of the following hardware and software
components.

Hardware Components:

• GSM Module.

• Microcontroller (Arduino UNO).

• Camera.

Software Components:

• Matlab.

• Arduino IDE

ISSN: 2005-4238 IJAST 3699


Copyright ⓒ 2020 SERSC
International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp.3698-3706

Fig 1: Block diagram of traffic violation and detection.

The proposed system working can be explained for three different scenarios

2.1 Detection of the Helmet

The initial and basic step in helmet detection is by first detecting the moving Vehicle.
Rather than sending and processing an entire video and making it complicated, it starts
collecting the starting few frames of the vehicle which makes to process a video in a
much simpler way. The foreground detector needs a minimum number of video frames to
start the Gaussian mixture model [2] as foreground segmentation often involves
undesirable noise. Now, the system starts finding the bounding boxes of the component
involved with the moving vehicle. Usually, more blobs are detected as moving vehicles
such as pedestrians, animals, etc. hence, we remove the blobs which are less than n
number of pixels in this scenario its 150 pixels, therefore, only moving vehicles frames is
left. The blobs are still left with many empty spaces and gaps it is removed by using
morphological opening. Morphological opening [2] removes the noise and fill the gaps
which make the blob coherent and helps is detecting moving vehicle perfectly.

The next main step is the classification of vehicles. Classification is made by a machine
learning algorithm [2]. Vehicles are basically of two type's four-wheelers and two-
wheelers here, it deals only with two-wheelers for helmet detection. The system
undergone training for the classification of the vehicles, training images involved many
surrounding objects such as trees, footpath, etc. the images implies how a vehicle is seen
on roads. A variety of training images are fed through the use of image augmentation [2].
Even though it's not the most effective way of showcasing real-world moving objects still
it is more than enough to show the effectiveness of the machine learning algorithms to
check its feasibility. The images are converted into grayscale and the raw pixel values are
fed to the classifier.

The final step involves helmet detection, the same above approach has been used. The
trained images are fed to the detector by cropping the two-wheelers by only restricting it
to the cropped image heads of the two-wheeler. Using this technique it can detect the
helmet and classify it based on the training dataset.

ISSN: 2005-4238 IJAST 3700


Copyright ⓒ 2020 SERSC
International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp.3698-3706

Fig 2: Flow chart for the detection of a helmet.

2.2 Number Plate Detection

The initial step is to acquire the image from the image acquisition unit from the video
source. Then, the plate region is extracted which is the main step in this algorithm.
Captured plate region image is converted into binary [1] i.e. 0s and 1s by setting a
threshold value of 0 for all pixels in the captured plate region image with luminance less
than the set threshold value and 1 for remaining pixels [1]. Segmentation of plate region,
the characters in the license place are constituted i.e. separated into individual characters.
Image filtering is improved by removing the noise and unwanted gaps using
Morphological opening [2]. After the character's segmentation operation, horizontal and
vertical smearing is done for identifying the character regions and save the extracted
information by sending it to the database.

ISSN: 2005-4238 IJAST 3701


Copyright ⓒ 2020 SERSC
International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp.3698-3706

Fig 3. Flow chart of number plate detection.

2.3 Signal Jump Detection.

In this technique, the camera is acting as a sensor to detect the status of traffic light
whether it's a red light or not. The same camera is used for monitoring and capturing of
the vehicles. Once the red traffic light is turned ON, the image processing algorithm
detects and asks the camera to capture the images of vehicles that have crossed the signal.
These captured images will be saved in the database and do the same operation used in
the above number plate detection approach for detecting the licensed number plate.

This proposed system detects various violations like without helmet detection and signal
jump detection and sends the message to the vehicle owners registered number with the
licensed number plate

3.RESULTS

In this system, the camera is placed on the traffic signal and captures the images. The
captured images are processed into an image processing unit. Different algorithms are
used to detect the non-presence of helmet and signal jump which are the violations.

Figure 4 explains that the number plate image is acquired from a video source. The region
of number plate is extracted and given as an input image to the simulator. The characters
in the number plate are separated into individual characters.

ISSN: 2005-4238 IJAST 3702


Copyright ⓒ 2020 SERSC
International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp.3698-3706

Fig 4: Simulated output of captured number plate input image.

Figure 5: Simulated output of extracted number plate characters.

Figure 5 explains the character segmentation operation, horizontal and vertical smearing
is done for identifying the character regions and save the extracted information by sending
it to the database.

The algorithm has been tested and the results have been found for various number plates
and different kinds of violations such as helmet and signal jump violation.

ISSN: 2005-4238 IJAST 3703


Copyright ⓒ 2020 SERSC
International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp.3698-3706

Figure 6 explains how helmet detection is made based on the above approach. The trained
images are fed to the detector by cropping the two-wheeler by only restricting it to the
cropped image heads of the two-wheeler. Using this technique it can detect the helmet and
classify it based on the training dataset.

Figure 6: Simulation results of a Helmet detection

Fig 7: Simulation of signal jump detection demo

Figure 7 explains about signal jump detection. The first part of the above figure tells us
that the car is behind the red light. Check still in the above figure is used to check the path
of the signal, the check love stream checks the live stream of the camera recordings when
the car passes the zebra cross line when the traffic red light is on it captures the number
plate and processes it to the database.

ISSN: 2005-4238 IJAST 3704


Copyright ⓒ 2020 SERSC
International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp.3698-3706

Once the algorithm gives a result as a violation has been made. The GSM Module sends
the message to the registered number of the licensed number plate of the violated vehicle.
The violated vehicle's owner receives a message about the vehicle number, violated the
rule and the penalized amount.

In this way, the algorithm was tested for different violations and the message report was
delivered successfully.

4.APPLICATIONS

A. Social relevance: The application built is mainly used for social relevance where
the detection of violation of traffic rules and sending an alert message which can be used
for educating road safety and alert then about the violation by making them pay fines
without the need of the human power

B. Tourism Platform: It is used to count to the number of heads of the traveling


passengers during the tourism entry and exit.

5.CONCLUSION

As discussed many lives have been lost due to road accidents by violating the traffics
rules and different problems faced by the policeman to maintain a database about the
violated information. This system provides a solution for all these problems in a much
simpler and easier way. The system developed was able to satisfy the 90 percent of our
required problem purpose. There are still unidentifiable violations like seat belt violation
etc. which is our future work of this system. The number plate was decoded and was able
to send the message to the vehicle owners.

In this way, the system can educate and give alert to the traffic rules violators by giving
caution to them telling you have violated the rules and the vehicle has been penalized with
the fine without the need for human power. The efficiency and performance of the system
are better which can survive for longer duration without getting failed.

REFERENCES

[1]. R Shreyas, Pradeep Kumar B.V, Adithya H.B, Padmaja B, Sunil M.P, "Dynamic
Traffic Rule Violation Monitoring System Using Automatic Number Plate Recognition
with SMS Feedback", 2017 2nd International Conference on Telecommunication and
Networks (TEL-NET 2017).

[2]. Dharma Raj KC, AphinyaChairat, VasanTimtong, Matthew N. Dailey,


MongkolEkpanyapong, "Helmet Violation Processing Using Deep Learning ", 2018 IEEE
ISBN 978-1-5386-2615-3/18.

[3]. Kunal Dahiya, Dinesh Singh, C. Krishna Mohan, "Automatic Detection of Bike-
riders without Helmet using Surveillance Videos in Real-time", 2016 International joint
Conference on Neural Networks (IJCNN).

ISSN: 2005-4238 IJAST 3705


Copyright ⓒ 2020 SERSC
International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp.3698-3706

[4]. XiaolingWang, Li-MinMeng, Biaobiao Zhang, Junjie Lu, K.-L.Du, "A Video-
based Traffic Violation Detection system", 2013 International Conference on Mechatronic
Schemes, Electric Engineering And Computer(MEC) Dec 20-22,2013, Shenyang China

[5]. Xiaopeng Ji, Zhiqiang Wei, Lei Huang, Yun Gao, "Violation Vehicle Automated
Snap And Road Congestion Detection", Proceedings of CCIS2016, Ocean University of
China, ISBN 16:978-1-5090-1256-5.

[6]. RhenAnjeromeBedruz, Aaron Christian P. Uy, Ana Riza Quiros, Robert


KerwinBillones, Edwin Sybingco, ArgelBandala, Elmer P. Dadios, "A Robotic Model
Approach of an Automated Traffic Violation Detection System with Apprehension",
Vol.2018 ISBN 18: 978-1-5386-7767-4/18.

[7]. Siddharth Tripathi, Uthsav Shetty, Asif Hasnain, Rohini Hallikar, "Cloud-Based
Intelligent Traffic System to Implement Traffic Rules Violation Detection and Accident
Detection Units", Proceedings of the Third International Conference on Trends in
Electronics and Informatics (ICOEI 2019) IEEE Xplore Part Number: CFP19J32-ART;
ISBN: 978-1-5386-9439-8.

[8]. Ali Sentas, Seda Kul, Ahmet Sayar, "Real-Time Traffic Rules Infringing
Determination Over the Video Stream: Wrong Way and Clearway Violation Detection".

[9]. Jingye Wang, Zhihua Chen, Ping Li, Bin Sheng, Ruoxi Chen, "Real-Time Non-
Motor Vehicle Violation Detection in Traffic Scenes", Vol.2019 ISBN:978-1-5386-8500-
6

[10]. MukreminOzkul, IlirCapuni, "Police-less multi-party traffic violation detection


and reporting system with privacy preservation", IET Intell. Transp. Syst., 2018, Vol. 12
Iss. 5, pp. 351-358.

[11]. Chaithra B., Karthik K.V., Ramkishore D., and Sandeep R., "Monitoring Traffic
Signal Violations using ANPR and GSM", International Conference on Current Trends in
Computer, Electrical, Electronics and Communication (ICCTCEEC-2017).

[12]. Julius T. Sese, Ramon G. Garcia, Ervin Jay V. Galvan, Joseph V. Rodriguez,
"Cloud-Based Detection and Real-Time Communication of Number Coding Violators
through character Recognition and Image Processing", 2019 IEEE 8th Global Conference
on Consumer Electronics (GCCE).

[13]. S. P. Mani Raj, Rupa. B, Potru Sarada Sravanthi, Geetha Sushma. K, "Smart and
Digitalized Traffic Rules Monitoring System", Proceedings of the International
Conference on Communication and Electronics Systems (ICCES 2018) IEEE Xplore Part
Number: CFP18AWO-ART; ISBN:978-1-5386-4765-3.

[14]. Ibtissam SLIMANI, Abdelmoghit ZAARANE, Abdellatif HAMDOUN and Issam


ATOUF, "Vehicle License Plate Localization and Recognition System for Intelligent
Transportation Applications", 2019 6th International Conference on Control, Decision
and Information Technologies (CoDIT'19).

ISSN: 2005-4238 IJAST 3706


Copyright ⓒ 2020 SERSC

You might also like