The Study Of Traffic Parameters Detection Technologies On Video

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 of traffic in the round. The methods of traffic parameters detection commonly used have many kinds of ways, but the video-based vehicle detection system has developed quickly in recent years, since this approach has many outstanding 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. Vehicles are drove by the roadway, and the vehicle detection is based on the traffic image disposal and analysis of each roadway, where sets a detecting region respectively. We can detect the traffic flow parameters through getting the characters of the image in the region, the trait of this method deal with little image area only, in doing that, we can decrease the operation capacity and computation load.This paper is take the traffic video on the highway as an example. And we have studied in depth in the area of image pretreatment technologies including background capture,image segmentation and image get rid of noises. On the basis of this, we have also designed the algorithms of traffic parameters detection by dragging the characteristic in the detecting region. The algorithms include vehicles counting and the velocity of vehicles. This paper mainly focuses on the following contents::(1)Studied and analyzed the process of dragging the moving objects in the traffic image on video, and it drag the moving objects by the background difference method this paper;(2)According to the characters of freeway, detecting the roadway structure with Hough transformation to get prepare for the calculating of the traffic parameters on each roadway;(3)Analyzed the image characteristics in the detecting region which are changed with the time, and judged the states of the vehicles;(4)Designed the counting algorithms of volume and velocity of traffic flow according to the location of vehicles.