Research And Application Of Road Identification And Traffic Parameter Statistics Methods Based On Video

With the development of society,traffic problems such as traffic jams and traffic safety have become increasingly prominent.To solve these traffic problems,establishing an intelligent transportation system is an effective method.The statistics of traffic parameters is one of the important contents of the intelligent traffic system.The real-time and accuracy of the traffic parameters will directly affect the performance of the intelligent traffic system.The video-based statistical method of traffic parameters is widely concerned with features such as high intelligence,wide application scope,low cost,and easy operation.It is an important research direction of traffic parameter statistics.At present,the statistical method of traffic parameters based on low-altitude fixed camera video is better in accuracy and real-time,and the traffic parameter statistics based on high-altitude fixed camera or drone video is affected by the interference of non-road areas in the image and the image motion.Because of the poor accuracy and real-time performance,it is of practical significance to carry out research on traffic parameter statistics based on high-altitude video images.This article attempts to automatically statistic the traffic parameters(Traffic flow,traffic density,driving speed)on the detected road to solve the problem of inaccurate and inaccurate statistics of parameters due to non-road region interference and image motion.The paper focuses on road detection technology and statistical methods for traffic parameters.The research content is as follows:(1)For the problem of non-road area interference,a road recognition algorithm based on edge detection is proposed based on the analysis of several existing road identification methods.The algorithm first uses Canny operator to carry out edge detection and enhancement on video images containing roads,and then uses Hough transform to match the detected edge shape features.Finally,by analyzing and calculating the edge features of the road,the automatic recognition of the road is realized.With segmentation.Experimental results show that this method has high recognition rate for road images in different backgrounds.(2)Aiming at the problems of misjudgment and real-time updating of background modeling in the detection of moving targets,based on researching and analyzing severalexisting methods of moving targets detection,a vibe algorithm based moving target detection method is proposed.The method first uses a single video frame to complete the initialization of the background model,and then compares the pixels to be classified with the corresponding background model,classifies the pixels into foreground pixels or background pixels,and finally adopts a “first-in,first-out” approach to update the background model.The accuracy of the target detection.Experimental results show that the algorithm has a good effect on the detection of moving vehicles.(3)The principle of Camshift target tracking algorithm is studied,and a vehicle speed calculation method based on target tracking is given.This method first uses the Vibe algorithm to obtain the foreground image of the moving vehicle in the image,then uses the Camshift tracking algorithm to perform matching tracking on the detected moving vehicle,obtains the position information of the moving vehicle after a certain interval frame,and finally obtains two frames of images according to the obtained image.The coordinates of the position of the middle-moving vehicle,the two central coordinate distances of the target vehicle in the actual road position are calculated through the coordinate transformation model,and the vehicle speed is calculated using the speed model.Experiments show that the given method has certain effectiveness.(4)Based on the analysis of video-based vehicle speed detection principle and existing detection methods,combined with the existing virtual coil method,a detection algorithm based on gray-scale matching is presented.The algorithm first uses the detected road and vehicle to automatically set the virtual coil,and then matches whether or not it is the same car by detecting the gray changes when the running vehicle passes through the bottom edges of the virtual coil.Finally,the vehicle speed is analyzed and calculated.The experimental results show that the accuracy of vehicle speed detection algorithm based on gray matching is higher,and its accuracy can reach 95.16%.(5)According to the system requirements and software architecture,using the research results of this paper,a video-based road identification and traffic parameter statistics system was designed and implemented.