Scribd
Upload
385 views

Automatic Traffic Signal Control Using Image Processing

This paper focus on the necessity of intelligent traffic system and the peculiar way of implementation with embedded system tools. Here it is implemented using an object counting methods and detection of emergency vehicles simultaneously thereby control the traffic signals based on the priority outcome. Methods/Statistical analysis: Presently various system do provide a cost effective solution, but the rate of successful operation is bad. Inductive loop detectors installed under surface, this fails in case of poor road condition and so. In the other system, the density is found by installing IR detectors along the side of the lane. These systems operate less efficiently where they are not applicable in the real time process. This method uses image acquisition method using a real time live video stream and the algorithm is processed using MATLAB. Results/Findings: The results obtained with the prototype are much encouraging and the system does help the present traffic control system to be more efficient. Conclusion/Application: Computer Vision being one of the most researched are for the future technologies, this system will add gain to this sector with efficient operation replacing the current primitive timer traffic control system. This helps the emergency casualties to be attended quickly without panic of traffic congestion. www.indjst.org/index.php/indjst/article/download/64622/57928

Uploaded by

MuraliShiva
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
385 views

Automatic Traffic Signal Control Using Image Processing

This paper focus on the necessity of intelligent traffic system and the peculiar way of implementation with embedded system tools. Here it is implemented using an object counting methods and detection of emergency vehicles simultaneously thereby control the traffic signals based on the priority outcome. Methods/Statistical analysis: Presently various system do provide a cost effective solution, but the rate of successful operation is bad. Inductive loop detectors installed under surface, this fails in case of poor road condition and so. In the other system, the density is found by installing IR detectors along the side of the lane. These systems operate less efficiently where they are not applicable in the real time process. This method uses image acquisition method using a real time live video stream and the algorithm is processed using MATLAB. Results/Findings: The results obtained with the prototype are much encouraging and the system does help the present traffic control system to be more efficient. Conclusion/Application: Computer Vision being one of the most researched are for the future technologies, this system will add gain to this sector with efficient operation replacing the current primitive timer traffic control system. This helps the emergency casualties to be attended quickly without panic of traffic congestion. www.indjst.org/index.php/indjst/article/download/64622/57928

Uploaded by

MuraliShiva
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Indian Journal of Science and Technology, Vol 8(16), DOI: 10.

17485/ijst/2015/v8i16/64622, July 2015

ISSN (Print) : 0974-6846


ISSN (Online) : 0974-5645

Smart Control of Traffic Signal System using Image


Processing
V. Parthasarathi1, M. Surya1, B. Akshay1, K. Murali Siva1 and Shriram K. Vasudevan2*
1

Department of Electrical and Electronics Engineering, Amrita School of Engineering, Amrita Vishwa Vidyapeetham
(University), Coimbatore - 641112, Tamil Nadu, India
2
Department of Computer Science and Engineering, Amrita School of Engineering, Amrita Vishwa Vidyapeetham
(University), Coimbatore, India; Kv_shriram@cb.amrita.edu

Abstract
Background/Objectives: This paper focus on the necessity of intelligent traffic system and the peculiar way of
implementation with embedded system tools. Here it is implemented using an object counting methods and detection of
emergency vehicles simultaneously thereby control the traffic signals based on the priority outcome. Methods/Statistical
analysis: Presently various system do provide a cost effective solution, but the rate of successful operation is bad. Inductive
loop detectors installed under surface, this fails in case of poor road condition and so. In the other system, the density
is found by installing IR detectors along the side of the lane. These systems operate less efficiently where they are not
applicable in the real time process. This method uses image acquisition method using a real time live video stream and the
algorithm is processed using MATLAB. Results/Findings: The results obtained with the prototype are much encouraging
and the system does help the present traffic control system to be more efficient. Conclusion/Application: Computer
Vision being one of the most researched are for the future technologies, this system will add gain to this sector with
efficient operation replacing the current primitive timer traffic control system. This helps the emergency casualties to be
attended quickly without panic of traffic congestion.

Keywords:

1. Introduction
Road transport is one of the primitive modes of transport
in many parts of the world today. The number of vehicles
using the road is increasing exponentially every day. Due
to this reason, traffic congestion in urban areas is becoming unavoidable these days. Inefficient management
of traffic causes wastage of invaluable time, pollution,
wastage of fuel, cost of transportation and stress to drivers, etc.1 but more importantly emergency vehicles like
ambulance get stuck in traffic. Our research is on density
based traffic control with priority to emergency vehicles
like ambulance and fire brigade. So, it is very much necessary to design a system to avoid the above casualties

*Author for correspondence

thus preventing accidents, collisions, and traffic jams2.


The common reason for traffic congestion is due to poor
traffic prioritization, where there are situations some lane
have less traffic than the other and the equal green signal
duration for both affect the wastage of resources and drivers are stressed.

2. Proposed System
The aim of this project is to control traffic signals with
the help of surveillance camera present at the junction
points. The frames of the traffics obtained from the camera through continuous video processing. To calculate
the density, an image from the camera is used to calcu-

Smart Control of Traffic Signal System using Image Processing

late the number of vehicles in each lane. According to the


number of vehicles in each lane, the time for respective
green signal is given which varies time to time. If there
are same numbers of vehicles in the lane, the signal will
follow the basic timer circuit. But when an emergency
vehicle such as an ambulance is detected, priority is given
for latter lane. The time for which each lane will be green
is shown in the display of the individual down counter.
In case, if an ambulance is detected, the current green
lane will become red and the counter display will show
an ambulance symbol, after a few seconds the lane having the ambulance will be allowed. Incase if there are two
ambulance detected in the junction, the ambulance which
is nearer to the signal get the priority first. Successful
implementation of our research will result in faster clearance of traffic and improvement in the transportation of
emergency vehicles.

2.1 Flow of Algorithm

2.2 Flow Chart Description


Figure 1 is the flow chart which depicts the intelligent
traffic management system using image processing3.
The camera which is mounted on a height gives the real
time traffic monitoring of the road junction2. Frames are
obtained periodically and are sent to the PC for image
processing. The traffic density estimation & ambulance
detection is done through image processing. If no ambulance is found then the traffic density is estimated &
signal is prioritized accordingly by the MCU & the signal
is toggled. But, if an ambulance is detected then signal
priority is given to that lane & if multiple ambulances in
different lanes are detected then priority is given to the
ambulance which is closer to the signal & the signal is
toggled accordingly.
The components used in this project are as follows:

2.2.1 Hardware Model


In this project the hardware used are:
USB web camera: To capture images,
PC: For all the image processing work,
MCU: Arduino board for signal prioritizing,

Toy cars for the prototype of a road junction and
traffic model.

2.2.2 Software Model


In this prototype modeling circuit, the density and ambulance detection is done using MATLAB and the traffic
signal is controlled by using MSP430 microcontroller.

3.Image Processing Steps


Involved for Vehicle Count and
Ambulance Detection
3.1Ambulance Detection
3.1.1 Input of Frames from Camera
The initial step is to obtain the frames of the road from the
live streaming of the camera. An image of the road without vehicles is taken as the reference frame (Figure 2 (a)).
Then images of the road with vehicles are taken periodically as the current frames. (Figure 2 (b), 2 (c))

3.1.2 Cropping of the Image to Obtain Required


Details
Figure 1. Flow chart of the proposed model.

Vol 8 (16) | July 2015 | www.indjst.org

The next step is to focus only on the region of interest.

Indian Journal of Science and Technology

V. Parthasarathi, M. Surya, B. Akshay, K. Murali Siva and Shriram K. Vasudevan

(a)

(c)

(b)

Figure 2. (a) Reference Frame, (b) Current Frame, (c) Current frame with ambulance.

(a)

(b)

(c)

Figure 3. (a) Cropping of background frame, (b) Cropping of current frame 1, (c) Cropping of current frame 2.

This is done by cropping and all the unwanted regions are


eliminated. The lanes of the road are our region of interest hence it is cropped. The reference frame of the lanes
(Figure 3(a)) and current frame (Figure 3(b)) and current
frame with ambulance (Figure 3(c)) are cropped.

lance with a standardized siren in our prototype is shown


in (Figure 4).

3.1.3 Detection of Ambulance


To detect the ambulance on a lane as shown in Figure 5
(a), its siren is being detected. The red color (Figure 5 (b),
5 (c)) & blue color (Figure 5 (d), 5 (e)) in the image is
obtained using image segmentation based on color4. Its
respective centroids are found out and distance between
each red & blue pair is calculated. If the distance between
them is lesser than a given predefined threshold distance,
and if both the centroids lie on the same vehicle then the
corresponding red & blue color is coming from the siren
and the vehicle is found to be an ambulance. The ambu-

Vol 8 (16) | July 2015 | www.indjst.org

Figure 4. Ambulance with red & blue colored siren in


prototype.

Indian Journal of Science and Technology

Smart Control of Traffic Signal System using Image Processing

(a)

(b)

(c)

(d)

(e)

Figure 5. (a) Cropped Image of Lane 1, (b) Red Color, (c) Binary image of Red objects, (d) Blue Color, (e) Binary image of
blue object.

3.2 Traffic Density Estimation


3.2.1 Detection of Vehicles
The next step is to detect the number of vehicles on the
road. To obtain this the current frame and the Background
frame are converted to gray scale (Figure 6 (a)) and the
images are compared and subtracted to obtain presence
of objects on the road3. Morphological operations such
as dilate & erode are carried out to remove the additional

(a)

noise in the image and this image is further enhanced and


it is converted to binary image (Figure 6 (b)). This image
is then filtered using Gaussian filter and to obtain only the
vehicles on the road.

3.2.2 Count of Vehicles


The next step is to count the number of vehicles present
on the roads3. To achieve these sets of connecting pixels

(b)

Figure 6. (a) Image after background subtraction (Grayscale), (b) Image after Enhancement (Binary Image).

Vol 8 (16) | July 2015 | www.indjst.org

Indian Journal of Science and Technology

V. Parthasarathi, M. Surya, B. Akshay, K. Murali Siva and Shriram K. Vasudevan

of our algorithm is cost effective. Extra hardware such as


sound sensors or RFID tags can be eliminated.

5. Future Work
The accuracy of this work can be improvised further
by doing thermal image processing5,6. Thermal image
processing is effective even during extreme weather conditions such as, mist or fog. Secondly, Cloud computing
can be done for the road data analysis.

6.References

Figure 7. The bounding of connecting pixels & obtaining


the vehicle count.

are labeled which are the vehicles & they are marked with
a bounding box (Figure 7).

4. Results and Conclusion


In this project we have successfully made the prototype
for real time image processing for smart automation of
traffic signal system for density estimation and emergency
vehicle detection such as ambulance4. Our model detects
the ambulance by detecting its siren. This is achieved
through image segmentation based on red & blue color of
the siren. The traffic density on each lane is also estimated
and the traffic signal is prioritized accordingly. The usage

Vol 8 (16) | July 2015 | www.indjst.org

1. Abbas N, Tayyab M, Qadri MT. Real Time traffic density


count using image processing. International Journal of
Computer Applications. 2013; 83(9):169.
2. 
Dharani SJ, Anitha V. Traffic density count by optical
flow algorithm using image processing. Second National
Conference on Trends in Automotive Parts Systems and
Applications (TAPSA-2014). India International Journal
of Innovative Research in Science, Engineering and
Technology. 2014 Apr; 3(2):5017.
3. Aishwarya S, Manikandan P. Real time traffic light control
system using image processing. IOSR-JECE. 15.
4. Dangi V, Parab A, Pawar K, Rathod SS. Image processing
based intelligent traffic controller. UARJ. 2012; 1(1):16.
5. Kannegulla A, Reddy AS, Sudhir KVRS, Singh S. Thermal
Imaging system for Precise Traffic Control and Surveillance.
International Journal of Scientific & Engineering Research.
2013 Nov; 4(11):4647.
6. Available from: http://www.flir.com/uploadedfiles/Eurasia/
MMC/Appl_Stories/AS_0049_EN.pdf. Accessed on 2015
Jan 5.

Indian Journal of Science and Technology

You might also like