Font Size: a A A

Development Of Traffic Information Video Capture System For Multi-environment Condition And Multi-vehicle Type

Posted on:2015-01-21Degree:MasterType:Thesis
Country:ChinaCandidate:W D WangFull Text:PDF
GTID:2272330422485670Subject:Traffic Information Engineering & Control
Abstract/Summary:PDF Full Text Request
The accurate data acquisition of the traffic flow and the study of traffic flow theory thatadapt to the situation of our country has been the key points to develop the ITS(IntelligentTransport System),relieve the traffic pressure, cut down traffic accident rate, and improvethe traffic capacity of the road. With the rapid development of computer technology, videoanalyze technology becomes very popular among the researchers. This paper designs a setof multi-vehicle video traffic flow information collection system, based on many trafficflow information collection methods, and develops a related software based on VS2010environment and Opencv(Open Source Computer Vision Library).This paper focuses on two aspects of researches. One is to develop the accurate andreliable information acquisition software. The other is to analyze and validate traffic flowtheory utilize the test results. This paper introduces several common methods of trafficinformation collection at first. Then elaborates the advantage and current situation of videotraffic flow information collection, analyzes the basic characteristic of traffic flow, discussesthe relationship of basic traffic parameters, and introduces few traffic flow probabilitystatistical models and traffic flow distribution characteristics of time and space on the basisof understanding parameters selection criteria.The key technology used in the system includes extracting the colored background,updating adaptively, using background difference, morphology processing on the binaryimage and the method for camera calibration. Then, demonstrate the results of the softwareprocess and operation by combining with the specific video data, meanwhile, display asimple method to set system parameter and a way to save the test results.At last, the results show that the detection methods can accurately meet the needs oftraffic analysis by comparing the test results with radar detection results. On the basis of theabove, analyze the results of the data to discuss the rule of traffic parameters and explain thetraffic flow theory.
Keywords/Search Tags:multi-vehicle information collection system, video detection, backgrounddifference, vehicle tracking, VS2010, Opencv2.4.4
PDF Full Text Request
Related items