Throughout the past decade, vehicular networks have attracted a great deal of interest\nin various fields. The increasing number of vehicles has led to challenges in traffic regulation.\nVehicle-type detection is an important research topic that has found various applications in numerous\nfields. Its main purpose is to extract the different features of vehicles from videos or pictures captured\nby traffic surveillance so as to identify the types of vehicles, and then provide reference information\nfor traffic monitoring and control. In this paper, we propose a step-forward vehicle-detection and\n-classification method using a saliency map and the convolutional neural-network (CNN) technique.\nSpecifically, compressed-sensing (CS) theory is applied to generate the saliency map to label the\nvehicles in an image, and the CNN scheme is then used to classify them. We applied the concept of\nthe saliency map to search the image for target vehicles: this step is based on the use of the saliency\nmap to minimize redundant areas. CS was used to measure the image of interest and obtain its\nsaliency in the measurement domain. Because the data in the measurement domain are much smaller\nthan those in the pixel domain, saliency maps can be generated at a low computation cost and faster\nspeed. Then, based on the saliency map, we identified the target vehicles and classified them into\ndifferent types using the CNN. The experimental results show that our method is able to speed up\nthe window-calibrating stages of CNN-based image classification. Moreover, our proposed method\nhas better overall performance in vehicle-type detection compared with other methods. It has very\nbroad prospects for practical applications in vehicular networks.
Loading....