Real time vision system.pdf

Real time vision system.pdf

ID:34308208

大小:1.33 MB

页数:18页

时间:2019-03-04

Real time vision system.pdf_第1页
Real time vision system.pdf_第2页
Real time vision system.pdf_第3页
Real time vision system.pdf_第4页
Real time vision system.pdf_第5页
资源描述:

《Real time vision system.pdf》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、TRANSPORTATIONRESEARCHPARTCTransportationResearchPartC6(1998)271±288Areal-timecomputervisionsystemforvehicletrackingandtracsurveillanceBenjaminCoifmana,*,DavidBeymerb,1,PhilipMcLauchlanb,2,JitendraMalikb,3aInstituteofTransportationStudies,UniversityofCaliforn

2、ia,Berkeley,CA,94720,USAbDepartmentofElectricalEngineeringandComputerSciences,UniversityofCalifornia,Berkeley,CA94720-1776,USAReceived30March1998;receivedinrevisedform3December1998;accepted3December1998AbstractIncreasingcongestiononfreewaysandproblemsassociat

3、edwithexistingdetectorshavespawnedaninterestinnewvehicledetectiontechnologiessuchasvideoimageprocessing.Existingcommercialimageprocessingsystemsworkwellinfree-¯owingtrac,butthesystemshavedicultieswithcongestion,sha-dowsandlightingtransitions.Theseproblemsstem

4、fromvehiclespartiallyoccludingoneanotherandthefactthatvehiclesappeardierentlyundervariouslightingconditions.Wearedevelopingafeature-basedtrackingsystemfordetectingvehiclesunderthesechallengingconditions.Insteadoftrackingentirevehi-cles,vehiclefeaturesaretrack

5、edtomakethesystemrobusttopartialocclusion.Thesystemisfullyfunc-tionalunderchanginglightingconditionsbecausethemostsalientfeaturesatthegivenmomentaretracked.Afterthefeaturesexitthetrackingregion,theyaregroupedintodiscretevehiclesusingacommonmotionconstraint.Th

6、egroupsrepresentindividualvehicletrajectorieswhichcanbeusedtomeasuretra-ditionaltracparametersaswellasnewmetricssuitableforimprovedautomatedsurveillance.Thispaperdescribestheissuesassociatedwithfeaturebasedtracking,presentsthereal-timeimplementationofaprototy

7、pesystem,andtheperformanceofthesystemonalargedataset.#1999ElsevierScienceLtd.Allrightsreserved.Keywords:Tracsurveillance;Wide-areadetection;Vehicletracking;Videoimageprocessing;Machinevision*Correspondingauthor.Tel.:+1-510-845-5121;fax:+1-510-642-1246;e-mail:

8、zephyr@eecs.berkeley.edu1D.BeymerisnowwithSRIInternational,333RavenswoodAve,MenloPark,Ca94025.E-mail:beymer@-ai.sri.com2P.McLauchlanisnowwiththeUniversityofSurrey,SchoolofEE,ITandMath.,Gu

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。