Sie sind auf Seite 1von 4

International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)

ISSN: 0976-1353 Volume 14 Issue 2 APRIL 2015.

An efficient traffic control system implementation


using Image Matching
PRANAV PATIL MS. SAMEENA PATHAN
4th Semester, Dept of ECE Faculty Dept of ECE,
Jain College of Engineering, Belgaum Jain College of Engineering, Belgaum
ppranav618@gmail.com
sameena.pathan.k@gmail.com

Abstract: Traffic congestion is a condition on road networks that high the signal will remain green for more time. Secondly it
occurs as use magnifies, and is described by slower speeds, will also communicate with the adjacent junction signal. Both
longer trip times, and augmented conveyance queuing. When the signals will collectively manage the traffic depending on
traffic postulate is great enough that the interaction between the density.
vehicles slows the speed of the traffic pour, this results in some
With the advances in the technology of micro
congestion. As demand approaches the competency of a road (or
of the intersections along the road), extreme traffic jam sets in. electromechanical systems (MEMS) and development in
When vehicles are fully stopped for epoch of time, this Wireless communications. It is possible to design such
conversation is known as a traffic jam or traffic snarl-up. For system to overcome daily problems of traffic congestion. The
this, we must need an efficient traffic control system. Automatic simplest way for controlling a traffic light uses timer for each
traffic control and surveillance are important for road usage phase. Another way is to use electronic sensors in order to
and management. Timers for each stage are the simplest way to detect vehicles, and produce signal that cycles. We propose a
control the traffic. Another way is to use electronic sensors in system for controlling the traffic light by image processing.
order to find out vehicles, and produce signal. Here we suggest a The system will detect vehicles through images.
system that implement image processing algorithm in real time
Real time traffic light control.
traffic light control which will control the traffic light efficiently.
A web camera is placed in each stage of traffic light that will Automatic traffic monitoring.
capture the still images of the road where we want to control the Detection of traffic density.
traffic. Then those captured images are successively matched Setting up a Road network.
using image matching with a reference image which is an empty
road image. The traffic is governed according to percentage of Reduce traffic congestion.
matching.
II. LITERATURE SURVEY

I. INTRODUCTION In 1868 illuminated traffic signal was installed in


London near the Houses of Parliament. The device had
In recent years, traffic congestion has become a semaphore arms to signal drivers and pedestrians during the
significant problem .Traffic congestion and jams are one of day and it had red and green lenses illuminated by gas for
the main reasons for increasing transportation costs due to the viewing at night. The police officer would turn the signal 90
wasted time and extra fuel. Building new roads and lanes are degrees to position the semaphore arms and the colored
just not possible any longer, but building intelligence into the lenses towards traffic as needed. This was the first recorded
roads and lanes with advanced technology is certainly use of illuminated red and green colors to control traffic. In
possible. Hence, there is need for a better and efficient traffic 1912 Salt Lake City, UT police detective Lester Wire built a
congestion system. The existing control systems have to be two-color traffic signal that used electric illumination. In
used in a more intelligent way in order to increase the traffic 1914 the American Traffic Signal Company installed two-
throughput and decrease total travel times. Hence, reducing color signals in Cleveland, Ohio. In 1920 William Potts, a
waiting times before traffic lights can save our society Detroit Traffic Police Superintendent, designed the first 4-
billions of rupees annually. To make traffic light controllers way three-color traffic control device that is recognized as the
more intelligent, we exploit the emergence of novel basis for the modern traffic signal. In India, Mumbai`s traffic
technologies such as communication networks and Image system is controlled by Real Time Intelligent Urban Traffic
processing by regulating the traffic demand at each Management Technology.
intersection in the network. The goal is to avoid traffic
conflicts and shorten the queue length at a stop line. Our aim
is to develop the system at signals. This system will have
multifunctional operations. The system will firstly measure
the traffic density at different signals and accordingly change
the time delays for traffic lights the side at which the traffic is

138
International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 14 Issue 2 APRIL 2015.

After edge detection procedure both reference and


real time images are matched and traffic lights can
be controlled based on percentage of matching.
Table I. Types of traffic signal systems: If the matching is between 0 to 10% - green light is
Type System Country on for 90 seconds. If the matching is between 10
Fixed time TRANSYT UK to50% - green light is on for 60 seconds. If the
1.
systems. matching is between 50 to 70% - green light is on
2. Plan SCATS Australia for 30seconds. If the matching is between 70 to 90%
generation - green light is on for 20 seconds. If the matching is
systems. between 90 to 100% - red light is on for 60 seconds.
3. Traffic SCOOT UK
responsive A camera will be mounted on the each lane of the signal on a
centralized junction. CCD Cameras will feed the video to a
systems. COMPUTER. A MATLAB based digital image processing
will be done on the video frames to calculate the amount of
traffic. The process of digital image processing is done in
following steps:
III. METHODOLOGY APPLIED
The steps that are followed are as follows. 1) Image acquisition: One or several image sensor produces
Image acquisition digital image. The image may be 2D or 3D from depending
RGB to gray conversion on the type of sensor used. Image sensor types include light-
sensitive cameras, range sensors, tomography devices,
Image enhancement
ultrasonic cameras, etc.
Image matching using edge detection
2) Image Segmentation: In the analysis of the objects in
Phase 1- images it is essential to distinguish between the objects of
interest and the rest. This latter group is also referred to as
Initially image acquisition is done with the help of the background. The pixels in the image will be converted
web camera. into a series of 0s and ls. This can be done by getting the
First image of the road is captured, when there is no grayscale value of a pixel and comparing it to a certain pixel
traffic on the road. value. Segmentation lessens the area of concern and makes
This empty roads image is saved as reference image the system run faster.
at a particular location specified in the program.
3) Motion Tracking: Can be off-line or real-time. In offline,
RGB to gray conversion is done on the reference
the entire image sequence is stored. The tracking uses frame
image. differencing algorithm to generate a motion map
Now correction is done on the reference gray image .
to achieve image enhancement. 4) Background Subtraction: The process of finding the
Edge detection of this reference image is done transient objects by comparing the current frame to the
thereafter with the help of Prewitt edge detection background image.
operator.
5) Morphological Operations:
(i) Opening: Removes foreground details that are smaller
Phase 2-
than the structuring element. It is also used as a noise
Images of the road are captured.
removal operator;
RGB to gray conversion is done on the sequence of (ii) Closing: Removes background detailsholes, gaps, inside
captured images. corners, etc.that are smaller than the structuring element. It
Now correction is done on each of the captured gray is also used to link components (i.e., broken lines) that should
image to achieve image enhancement. be connected.
Edge detection of these real time images of the road
is now done with the help of prewitt edge detection 6) Image Correlation: The process of computing the
correlation for each Region of Interest (ROI) of the current
operator.
image to the ROI of the background image.
Real time image is captured and processed using image
Phase 3-
processing. Various operations are performs such as
subtraction operation, gray scale operation, binary operation,

139
International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 14 Issue 2 APRIL 2015.

opening operation, fill hole operation, erosion operation and Fig.2. Images for heavy traffic (a) Difference Image (b)
after that get total number of vehicles present on junction. Threshold BW (c) Colored-Labeled Objects
The final count is given to master microcontroller through
serial communication using RS232. The figure 3 indicates the visual appearance of image when
there is heavy traffic.

IV. RESULTS AND DISCUSSION


The simulation is performed in Matlab 7.2 version. The
figure 1 indicates the visual appearance of image when there
is no traffic.

(a) (b) (c) Fig.4. Circuit Simulation


Fig.1. Images for no traffic (a) Difference Image (b) Emergency Vehicle Detection:
Threshold BW (c) Colored-Labeled Objects The video is also analyzed for the detection of
emergency vehicles through their flashing red lights. By
The figure 2 indicates the visual appearance of image when specifying a threshold, we have isolated the areas with high
there is moderate traffic. intensity of red light and comparatively lesser intensity of
blue and green color. The resultant image is shown in Fig.
(b). As we can see, the headlights of the vehicle were also
detected, which led to an erroneous output. Hence the red
light must satisfy the additional condition of blinking. This is
achieved by taking account for the fact that the red light shall
appear in every third frame only. The other lights do not
appear in the image sequence with this frequency and hence
are eliminated. This leads to the conclusion of the presence of
an emergency vehicle as shown in (c). Our model was tested
for ambulance during various times of the day and was found
to be successful. In addition, the beacon can be identified
(a) (b) (c) even if the emergency vehicle is in an inclined position with
Fig.2. Images for moderate traffic (a) Difference Image (b) respect to the camera as shown in Fig(c).
Threshold BW (c) Colored-Labeled Objects.

(a) (b) (c)

(a)

140
International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 14 Issue 2 APRIL 2015.

REFERENCES

[1] A. Levin, D. Lischinski, and Y. Weiss, Colorization using


optimization, ACM Trans. Graph., vol. 23, no. 3, pp. 689694, Aug.
2004.
[2] Chein Pen Chaung and G X Chen, A lossless color image
compression algorithm with Adaptive Arithmetic Coding, Intelligent
Systems International Conference, 2012.
[3] G Chenchu Krishnaiah, T Jayachandraprakash, M N Giriprased.
Efficient Image Compression Algorithms using Evolved wavelets,
International Journal of systems and Technologies, Volume 4. No 2.
[4] K.Sowmyan, A.Siddarth, D.Menaka, A Novel approach to Image
compression of Color Images by Plane Reduction Techniques, World
Academy of Science, Engineering and Technology 57 2011.
[5] M.A. Ansari and R.S. Anand, Recent trends in image compression and
its application in telemedicine and teleconsultation, XXXII, NSC
(b) 2008, December 17-19, 2008.
[6] Mr. S.V.Viraktamath, Dr. Girish V. Attimard. Impact of Quantization
Matrix on the Performance of JPEG International Journal of Future
Generation Communication and Networking Vol. 4, No. 3, September,
2011.
[7] Nageswara Roa Thota, Srinivasa Kumar Devireddy. Image
Compression using DCT, Georgian Electronic Scientific Journal CS
and Telecommunications 2008 Volume No.3. J. U. Duncombe,
Infrared navigation-Part I An assessment of feasibility, IEEE Trans.
Electron Devices, vol. ED-11, pp. 34-39, Jan. 1959.
[8] Sukho Lee, Sang-Wook Park, Paul Oh, and Moon Gi Kang,
Colorization based Image Compression using Optimization, IEEE
transactions on image processing, VOL. 22, NO. 7, JULY 2013.
[9] Ueno, T, Yokohama, Yoshida, T. Ikehara, M. Color Image Coding
Based in Colorization, APSIPA, Asia- Pacific, IEEE 3-6 December
2012.
[10] Xin Chen, Sam Kwong, and Ju-fu Feng, A New Compression Scheme
for Color-Quantized Images, IEEE transactions on circuits and
systems for video technology, Vol. 12, No. 10, October 2002.
(c)
Fig. 5. (a) Image of a vehicle during diurnal (b) Detection of
all lights (c) Emergency vehicle detected.

Forming ANN (Artificial Neural Network):


The systems output is in form of Decimal value. This
value depends upon the no. of vehicle or the color density
detected. It may be difficult to add these value in a loop and
give them to a micro-controller for signal switching. Rather
ANN can be formed which will provide the output in 3
levels:
1. -1 for no/low traffic,
2. 0 for moderate traffic, &
3. +1 for heavy traffic.

The neurons can be so trained that they can count the no. of
vehicles or measure the color density and provide the output
in above mentioned levels. This will increase the accuracy of
the system and will be an easy task for signal switching.

V. CONCLUSION AND FUTURE SCOPE


The aim to detect the density of traffic is achieved in
this paper. The matlab algorithm detects the density of
traffic ie., either LOW, Moderate, or HIGH. In this paper
we have successfully achieved to change the states of
signals based on the traffic density. Matlab algorithm
successfully communicates with other signals and also
changes their timer values. Digital image processing is
also done succesffuly.

141

Das könnte Ihnen auch gefallen