Application Of Video Processing Technique In Intelligent Transportation System

The detection of traffic flow parameters on video plays a critical role in the intelligent transportation system (ITS); it enables the supervisor of the traffic to know the travel status in the round. The commonly used methods of traffic parameters detection have many kinds of ways, but the video-based vehicle detection system has developed quickly in recent years. Since this approach has many particular advantages which other methods don't have, it gradually becomes a research focus in the field of intelligent transportation system.Video-based traffic flow parameters detection system has several apparent advantages such as easy to see and install and low expense, thus it has represented the development direction of vehicles detector. At the crossroad of the city traffic, vehicles are drove along the roadway, and the vehicle detection is based on the traffic image disposal and analysis of each roadway. So we can firstly set detecting regions respectively and then detect the traffic flow parameters through each detecting region.This paper has taken the traffic video on the crossroad of the city as an example. The first step is studying the image pretreatment technologies including background capture,image segmentation and getting rid of image noises. On the basis of these, we have also designed an intelligent city traffic controlling model. The model includes two parts: the first part is intelligent control of the traffic lights on the basis of vehicles counting. and the second part is the illegal vehicles tracking. This paper mainly focuses on the following contents:(1)Studied and analyzed the process of extracting the moving objects in the traffic image on video, and it extracts the moving objects by the background difference method in this paper;(2)Analyzed the image characteristics in the detecting region which are changed with the time, and finished the detection of traffic flows.(3)An intelligent city traffic controlling model is proposed. Based on this model, we first obtain parameters of video traffic flows and the illegal traffic information in every direction of the crossroad by using video image processing technique. Then we implement intelligent control of the traffic lights and the illegal vehicles tracking.Finally, the intelligent city traffic controlling model has been designed by Visual C++ 6.0 and OpenCV, and the algorithms used in this paper almost have been realized on the PC, and it proved to be available.