Research Of Vehicle Flow Detection Technology Based On Video Image

Intelligent video surveillance of Intelligent transportation system is a part ofcomputer visual field,as a leading and emerging subject,it has received extensiveconcern and get a lot of research.Intelligent surveillance which processes under theunsupervised conditions split the video into image sequence, and then track the targetobject of the images, detect and monitor the behavior of target objects.At the sametime surveillance system observe the results of the analysis and determination,ultimately conduct warning if something is wrong. Intelligent monitoring due to itsunsupervised feature overcome many shortcomings of traditional monitoring systems,eliminating the need of using too many human resources.And at the same time,usingintelligent surveillance overcome the shortcoming of observation,therefore,the finalresults may relatively more accurate.Intelligent monitoring technology has beenwidely used in many of the high security requirements places,such as banks, militarybases,and high-speed road and other places.The primary means of intelligentmonitoring is the use of video technology to capture the scene,which has become theintelligent transportation system,a vital part,and the use of video technology can covera wide range of scenarios, you can also get a lot of information through it.Vehicle-specific methods studied in this paper is mainly through the videoprocessing,to obtain specific information of the target vehicle movement,and theshadow of the vehicle under different circumstances,be removed to exclude theinterference of the shadow vehicle counting and the target intersection.The processmainly consists of moving target detection,shadow removal and vehicle countingthese three parts.Motion target detection mainly through common backgroundmodeling method for video background modeling,then through the gap between theimage and background find target movement.Background modeling method weproposed in this paper using a histogram of the composite filter,the algorithmcompared to other algorithms in terms of computing easier,faster updates in real-timebackground.Shadow removal method uses a combination of shadow removal method,this method has good performance properties which can effectively identify the position of the shadow and remove it.The combination of morphological methods caneffectively remove empty.The vehicle counting section presents vehicle detectionmethod based on the coil which has achieved good results and it extended to arectangular area,and ultimately can effectively cross road for effective judgments.