| Abstrakt: |
Due to the extreme demand for an Intelligent Transportation System(ITS), there are numerous applications for detecting, monitoring, and numbering of moving vehicles on the road in real-time for surveillance and controlling the traffic. Computer vision-based techniques are more appropriate because they do not disrupt traffic during installation and are simple to modify. From pattern recognition to processing live football game action, computer vision exceeds human visual abilities in many areas. In this paper, the system was designed using Anaconda platform and OpenCV image development kits for vehicle counting, speed-estimation, and license plate detection. The chosen strategy finds foreground objects in a video sequence by using the background subtraction technique. Several computer vision techniques, such as image dispersion, frame differencing, and image thresholding, are used to count moving vehicles more accurately. We used SVC (4 cross-validations), segmentation, Connected Component Analysis (CCA) to detect the license plates. Also image segmentation,and image acquisition to detect the speed of a vehicle. Our system is inexpensive and portable, that makes it useful for civil, military, and government applications such as toll collection and highway monitoring. [ABSTRACT FROM AUTHOR] |
