The Research Of Vehicle Violation Discriminant Based On The Video And The Realization Of Wireless Transmission

In the face of increasingly serious traffic and traffic accidents, the video monitoring system of the field of intelligent traffic has been given more and more attention by scholars both at home and abroad. As computer vision, image processing, pattern recognition and artificial intelligence technology is applied to the intelligent video monitoring system, These techniques also gradually become focus and hot spot of the research. However, because of the condition of illegal turn decision is complicated, domestic now still rely on artificial methods for the recognition of the illegal turn. Therefore, the intelligent detection method of illegal turn has important research significance.Paper is based on the video vehicle tracking, distinguish abnormal behavior of the vehicle, the research and implementation of wireless transmission of video. Main job is to:1. The research of vehicle tracking algorithm. Paper first introduces the application in the field of tracking about Meanshift algorithm and kalman algorithm, then put forward a new algorithm combining kalman and Meanshift algorithm. The new algorithm is improved greatly in terms of tracking than Meanshift algorithm, when the foreground and background color is similar, it can track moving targets and still not lost, or when meeting more complicated scene it can also track moving targets. The classification of the multiple target tracking state is proposed,and then defined the emergence of a new target, the disappearance of the target, and target tracking, When tracking multi-target, the algorithm after defining is reduced in complexity.2. The research of discriminant method about Abnormal behavior of the vehicle. Paper’s main research is based on the traditional trajectory identification method. Based on the trajectory to discriminate the coordinates of the change, also is the change of vehicle coordinates to determine the behavior of the vehicle. Setting threshold value range about the coordinates change of turn left, turn right and go straight. Effect on the actual deviation, increasing the discriminant of frames in a row, then determine the behavior of the vehicle, finally identifying vehicles wheather have illegally turning or running red lights according to the vehicle lanes they belong to and traffic lights at intersections.3. Wireless video transmission based on the technology of wireless. Paper Mainly introduced the characteristics of UWB technology, making use of the NS simulation of UWB transmission and introduced the characteristics of the wireless bridge technology. Due to the limitations of transmission distance and technology of UWB, the transmission distance of UWB can only around10m, the corresponding product technology development is not yet mature. However, wireless bridges technology not only has wide bandwidth in terms of transmission, and transmission distance can reach kilometers. Finally choose wireless bridge solve the problems of the distance and the difficulty of transmission between the intersection of video switches. Then transmitting the video to the laboratory over a long distance by using the technology of wireless bridge. Paper set the video acquisition and transmission system in outdoor environment of simulation crossroads, the system can capture the video which is in400m away, then transmit the video back to the laboratory by using wireless bridge. Finally in view of the video image for vehicle detection, tracking, and illegal discrimination on the server based on vc++6.0platform by using OpenCV.