Font Size: a A A

Video-based Traffic Information Acquisition System

Posted on:2009-06-10Degree:MasterType:Thesis
Country:ChinaCandidate:S LinFull Text:PDF
GTID:2208360245956223Subject:Carrier Engineering
Abstract/Summary:PDF Full Text Request
The effective acquirement of traffic flow characteristic parameters has become the key point to the traffic flow theory and traffic micro simulation study,which currently are the mainstream and advanced means of the traffic basic theory study. And the characteristic parameters acquirement largely depends on the excellent traffic information collecting system.However,a majority of collecting systems currently only can provide such macro traffic parameters as traffic volume,density,occupancy, speed,header time,et al.Although they can meet the traffic control demand,there exists long distance for the traffic flow theory and traffic micro simulation study because collecting systems can not provide such traffic individual's micro parameters as operating trajectory,velocity and acceleration.As an important part of parameters collection,video detection technology has developed fast.It offers many virtues over the existing methods because it integrates such advanced technologies as image processing,computer vision and pattern recognition.And it can decrease the artificial impact on the collecting data and save material resources.This paper proposes a traffic information collecting method based on video detection and the software Video Track is developed in VC++ 6.0 with OpenCV.The major works are as follow:1,The system framework is built after the traffic flow parameters collecting technologies are comprehensively analyzed and concluded.2,The direct linear transformation algorithm(for short DLT)is presented for the traffic scene calibration and coordinate transformation,and the DLT algorithm is implemented with C++ language in VC++ 6.0 environment.3,The feature point optical flow tracking algorithm is used to track the motion objects such as vehicles,passengers in the traffic scene.The data are put into Excel automatically and achieves macro,so that it reaches the goals of the automatic collecting of traffic flow micro parameters such as velocity,acceleration and the trajectory.4,Finally,the traffic volume and speed detection algorithm based on virtual loop is presented and it is applied simply.The software developed in this paper is validated using true traffic scene video. The results indicate that the algorithms proposed are valid and the software VideoTrack has a bright application furore.
Keywords/Search Tags:traffic flow, traffic information collection, video detection, OpenCV
PDF Full Text Request
Related items