Vehicle tracking and speed estimation for traffic surveillance
Vehicle tracking is one of the critical applications of object detection and tracking. Traffic surveillance has become crucial in this day and age where the number of vehicles on the road has risen considerably. To preserve the safety of motorists, traffic law enforcement assign speed limits a...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English English |
| Published: |
2014
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/1524/1/24p%20KAIRIL%20FARIQ%20CHAIROL%20MOHD%20FEROZ.pdf http://eprints.uthm.edu.my/1524/2/KAIRIL%20FARIQ%20CHAIROL%20MOHD%20FEROZ%20COPYRIGHT%20DECLARATION.pdf http://eprints.uthm.edu.my/1524/3/KAIRIL%20FARIQ%20CHAIROL%20MOHD%20FEROZ%20WATERMARK.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Vehicle tracking is one of the critical applications of object detection and tracking.
Traffic surveillance has become crucial in this day and age where the number of
vehicles on the road has risen considerably. To preserve the safety of motorists,
traffic law enforcement assign speed limits at different locations throughout the
country. However, irresponsible motorists still exceed the speed limit since they
know it is unlikely that they will get caught. In this paper, a system is developed
which is capable of detecting moving vehicles in a video and display the vehicles
speed as it goes. Should a vehicle exceed the allowed speed limit, it will be displayed
in the video alongside the vehicle so that traffic law enforcers will be able to take
necessary action based on the displayed speed. The system uses Matlab/Simulink as
a simulation platform as it provides comprehensive tools for thresholding, filtering
and blob analysis. Optical flow was the image processing technique used to
determine the moving vehicles. A median filter was used to remove salt and pepper
noise from the thresholded image. Combinations of several morphological operations
were used to rectify whatever that is left. Blob analysis produces rectangles around
the moving objects. The centroid of the rectangle is used to determine the location of
each vehicle at a given frame. To make up for the absence of depth perception, the
camera’s height and angle from the road is fixed so that the rate of which a vehicle
approaches the camera can be determined. The results show that the system
successfully detects vehicles and displays its speed, though there is a relatively small
margin of error for the displayed speed. The displayed speed is set to only change
once every couple of frames so that it would be easier to see. |
|---|