资源描述：
《Intelligent multi-camera video surveillance A review智能多摄像机视频监控综述》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、PatternRecognitionLetters34(2013)3–19ContentslistsavailableatSciVerseScienceDirectPatternRecognitionLettersjournalhomepage:www.elsevier.com/locate/patrecIntelligentmulti-cameravideosurveillance:AreviewXiaogangWang⇑DepartmentofElectronicEngineering,TheChines
2、eUniversityofHongKong,Shatin,HongKongarticleinfoabstractArticlehistory:Intelligentmulti-cameravideosurveillanceisamultidisciplinaryfieldrelatedtocomputervision,patternAvailableonline20July2012recognition,signalprocessing,communication,embeddedcomputingandima
3、gesensors.ThispaperreviewstherecentdevelopmentofrelevanttechnologiesfromtheperspectivesofcomputervisionandKeywords:patternrecognition.Thecoveredtopicsincludemulti-cameracalibration,computingthetopologyofMulti-cameravideosurveillancecameranetworks,multi-came
4、ratracking,objectre-identification,multi-cameraactivityanalysisandMulti-cameracalibrationcooperativevideosurveillancebothwithactiveandstaticcameras.Detaileddescriptionsoftheirtechni-Topologyofcameranetworkscalchallengesandcomparisonofdifferentsolutionsarepro
5、vided.Itemphasizestheconnectionandinte-Multi-cameratrackinggrationofdifferentmodulesinvariousenvironmentsandapplicationscenarios.AccordingtothemostObjectre-identificationMulti-cameraactivityanalysisrecentworks,someproblemscanbejointlysolvedinordertoimproveth
6、eefficiencyandaccuracy.Withthefastdevelopmentofsurveillancesystems,thescalesandcomplexitiesofcameranetworksareincreasingandthemonitoredenvironmentsarebecomingmoreandmorecomplicatedandcrowded.Thispaperdiscusseshowtofacetheseemergingchallenges.Ó2012ElsevierB.V
7、.Allrightsreserved.1.Introductionvideosurveillancesystemshavebeendeveloped(Collinsetal.,2001;AghajanandCavallaro,2009;ValeraandVelastin,2004).ItIntelligentvideosurveillancehasbeenoneofthemostactiveisamultidisciplinaryfieldrelatedtocomputervision,patternrec-r
8、esearchareasincomputervision.Thegoalistoefficientlyextractognition,signalprocessing,communication,embeddedcomputingusefulinformationfromahugeamountofvideoscollectedbysur-andimagesensors.Thispaperreviews
