Intelligent Vehicle Recognition based on Wireless Sensor Network
Maha Mohamed Nabeel, Sherine M. Abd El-Kader · 2013
One of the main requirements of any intelligent transportation system is to be able to identify vehicles in the traffic. This paper presents an intelligent vehicle identification system used within a complete solution for a traffic monitoring system thatusesa novelwireless sensor network architecture to monitor traffic. A novel wireless sensor network architecture to monitor traffic is proposed where a visual sensor node captures images of the traffic and sends them to the traffic control center for processing. Also, this paper compares between three main localization and recognition algorithms. To locate the vehicle logo in the traffic image a symmetry detection algorithm is used to detect the inherent symmetry in vehicle frontal images. A fine location of the logo is identified using three different methods in the region marked by the symmetry line. After locating the logo three feature sets are extracted and presented to the classifier to correctly identify the type of the vehicle. The results of the localization and recognition algorithms show the efficiency of the presented system in identifying vehicle types with a recognition rate over 90%.