Vol 6 4 469 480 SHAFIE PDF
Vol 6 4 469 480 SHAFIE PDF
Vol 6 4 469 480 SHAFIE PDF
Abstract
Traffic signal light can be optimized using vehicle flow statistics obtained by
Smart Video Surveillance Software (SVSS). This research focuses on efficient
traffic control system by detecting and counting the vehicle numbers at various
times and locations. At present, one of the biggest problems in the main city in
any country is the traffic jam during office hour and office break hour.
Sometimes it can be seen that the traffic signal green light is still ON even
though there is no vehicle coming. Similarly, it is also observed that long
queues of vehicles are waiting even though the road is empty due to traffic
signal light selection without proper investigation on vehicle flow. This can be
handled by adjusting the vehicle passing time implementing by our developed
SVSS. A number of experiment results of vehicle flows are discussed in this
research graphically in order to test the feasibility of the developed system.
Finally, adoptive background model is proposed in SVSS in order to
successfully detect target objects such as motor bike, car, bus, etc.
Keywords: Vehicle detection, Motion detection, Traffic density estimation,
Traffic signal control.
1. Introduction
In recent years, video processing techniques have attracted researchers for vehicle
detection [1-6]. Neural networks have been widely used in traffic control system
[7-9]. Traffic incident detection model using neural networks has been developed
using traffic magnetic sensors [7]. Intelligent agent systems have been used in
order to control the traffic [8]. Hybrid computational intelligent techniques and
fuzzy neural networks have been applied in multi- agents in order to control the
traffic signals. They have reduced the average waiting time in the traffic. Traffic
469
470 A. A. Shafie et al.
Abbreviations
BS Background subtraction
ED Event database
FAR False acceptance rate
FRR False rejection rate
SVSS Smart video surveillance software
TSR Total success rate
TV Threshold value
VD Video database
flow prediction is achieved using time delay based neural networks [9]. In this
paper, we will use image processing technique for identification of vehicles and
traffic density by processing the traffic videos.
The pre-processing algorithm determines the contour of an object depending
on the application. Once the object is detected and located, its boundary can be
found by using edge detection and boundary following algorithms [10]. For object
detection and recognition in video surveillance, various approaches have been
proposed in the past years by many experts on this topic [11, 12].
Viitaniemi and Laaksonen [13] proposed a technique of moving object
detection in video analysis. They considered all the test image segments given by
a certain segmentation method. The segments have been ranked using the
PicSOM framework according to their similarity of target object segments in the
training images. They showed the average precision measurement for the test set
object detection. Hampapur et al. [14] proposed the different approach, called
salient motion detection, assumes that a scene will have many different types of
motion, of which some types are of interest from a surveillance perspective.
Fig. 2. Camera View Angles are shown in Details from Different Positions.
object. It is also noticeable that if the ratio falls under the range of 0.9<R<1.2 then
it will be saved as motor bike. Similarly, if the ratio is between 0.3 and 0.6
then the detected object will be saved as car and if greater than 0.6 then it will be
saved as bus.
Figure 8 highlights that the numbers of vehicles are decreasing with the
increment of time. Using this data traffic signal light can be control efficiently at
that place. The significant of the cloudy environment is to justify the accurate
vehicle detection rate under this condition in order to implement practically. The
special feature used in this case is the moving vehicle’s blob segmentation. The
image from edge filter is merged with the original image, producing image with
edge around motion area. Then this edge will be changed to color channel. This is
solely to emphasize the region where the motion actually occurs. Subsequently
this produces final image with colored blobs. The blob ratio has been calculated
by counting the pixel of Height/Width of the image.
The above discussion and results can be summarized as follows. The most
accurate result was found at an angle of 0 degree and 45 degree while camera
height was 0.5 m and 6 m respectively. The nearest result was at 15 degree angle
and 4 m height followed by 5 degree angle and 0.5 m height. On the other hand,
majority of the false detections occurred at 70 degree and 60 degree angles and a
height of 6 m from the ground. Besides this, there were false detections that
occurred at night and in cloudy environment due to occlusion.
Traffic flow control by SVSS in machine vision is the main contribution to the
existing system. The following future works will enhance the system to solve the
sophisticated security problems extensively:
• Night vision camera or infrared camera will help in recording the videos
even at night or at dark place. Collision and damage detections are
important areas that can benefit from such advance system.
• IP camera should be used to notify any danger or suspect acts to the user in
real-time alert system.
• High camera resolution should be selected where very clear images are necessary.
• Wireless camera should use in order to reduce wiring and maintenance cost.
References
1. Ji, X.; Wei, Z.; and Feng, Y. (2006). Effective vehicle detection techniques
for traffic surveillance systems. Journal of Visual Communication and Image
Representation, 17(3), 647-658.
2. Lozano, A.; Manfredi, G.; and Nieddu, L. (2009). An algorithm for the
recognition of levels of congestion in road traffic problems. Mathematics and
Computers in Simulation, 79(6), 1926-1934.
3. Zhou, J.; Gao, D.; and Zhang, D. (2007). Moving vehicle detection for
automatic traffic monitoring. IEEE Transactions on Vehicle Technology,
56(1), 51-59.
4. Niu, X. (2006). A semi-automatic framework for highway extraction and
vehicle detection based on a geometric deformable model. ISPRS Journal of
Photography and Remote Sensing, 61(3-4), 170-186.
5. Zhang, W.; Fang, X.Z.; and Yang, X. (2006). Moving vehicles segmentation
based on Bayesian framework for Gaussian motion model. Pattern
Recognition Letters, 27(1), 956-967.
6. Li, X.; Liu, Z.Q.; and Leung, K.M. (2002). Detection of vehicles from traffic
scenes using fuzzy integrals. Pattern Recognition, 35(4), 967-980.
7. Messai, N.; Thomas, P., T.; Lefebvre, D..; and Moudni, A.El. (2005). Neural
networks for local monitoring of traffic magnetic sensors. Control
Engineering Practice, 13(1), 67-80.
8. Srinivasan, D.; Choy, M.C.; and Cheu, R.L. (2006). Neural networks for real-
time traffic signal control. IEEE Transactions on Intelligent Transportation
Systems, 7(3), 261-272.
9. Baher, A.; Porwal, H.; and Recker, W. (1999). Short term freeway traffic
flow prediction using genetically optimized time-delay-based neural
networks. In Transportation Research Board 78 th Annual Meeting,
Washington, D.C.
10. Loncaric, S. (1998). A survey of shape analysis techniques. Pattern
Recognition, 31(8), 983-1001.
11. Yang, M.H.; Kriegman, D.J.; and Ahuja, N. (2002). Detecting faces in
images: a survey. IEEE transactions on Pattern Analysis and Machine
Intelligence (PAMI), 24(1), 34-58.
12. Hjelmas, E. (2001). Face detection: A survey. Computer Vision and Image
Understanding, 83(3), 236-274.
13. Viitaniemi, V.; and Laaksonen, J. (2006). Techniques for still image scene
classification and object detection. ICANN 2006, Lecture Notes in Computer
Science, 4132, 35-44.
14. Hampapur, A.; Brown, L.; Connell, J.; Ekin, A.; Haas, N.; Lu, M.; Merkl, H.;
and Pankanti, S. (2005). Smart video surveillance: exploring the concept of
multiscale spatiotemporal tracking. IEEE Signal Processing Magazine, 22(2),
38-51.