资源描述：
《Crowd Abnormal Behavior Recognition in Intelligent Surveillance System》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、AppliedMechanicsandMaterialsVols.263-266(2013)pp2592-2596Online:2012-12-27©(2013)TransTechPublications,Switzerlanddoi:10.4028/www.scientific.net/AMM.263-266.2592CrowdAbnormalBehaviorRecognitioninIntelligentSurveillanceSystem1,a2,b3,cYingZhao,TianfeiZhouandShaoqianWei1No.5Waig
2、uanxiejie,AndingmenWai,ChaoyangDistrict,Beijing,China2No.5SouthZhongguancunStreet,HaidianDistrict,Beijing,China3No.5Waiguanxiejie,AndingmenWai,ChaoyangDistrict,Beijing,Chinaabcsftzhaoying@hotmail.com,sfttianfei@163.com,sftshaoqian@163.comKeywords:CrowdBehavior•CrowdModeling•C
3、rowdAbnormalBehaviorRecognition•IntelligentSurveillanceSystemAbstract.Crowdabnormalbehaviorrecognitionisessentialforintelligentvisualsurveillanceinpublicplacestoensurethesafetyofthepublic.Thisisachallengingworkbecausecrowdbehaviorsarecomplexwhichareinfluencedbyvariousfactors.
4、Thispaperdividedthesefactorsintothreecategories:physicalfactors,socialfactorsandpsychologicalfactors.Thenanoverviewaboutcrowdbehaviormodelingapproacheswasgiven.Afterthat,thepaperdescribedandanalyzedsomeinfluentialexistingalgorithmsincrowdabnormalbehaviorrecognitionfromtheview
5、pointofbehavioralfactorstheyused.Finally,thepaperdiscussedthefutureresearchdirectionsinthisareaandsomeresearchproposalsweregiven.IntroductionVideosurveillancesystemsareverycommoninpublicplacesnowadays,buttheyaremainlyforrecording.Alloftheimageunderstandingandriskdetectionisle
6、fttohumansecuritypersonnel.However,thistypeofobservationtaskisnotwellsuitedtohumansbecauseitrequirescarefulconcentrationoverlongperiodsoftime.Therefore,thereisclearmotivationtodevelopautomated,intelligent,vision-basedsurveillancesystemswhichcanaidahumanuserintheprocessofabnor
7、malactivitiesdetectionandalarmasautomaticallyaspossible.Becauseofthehighlevelofdegenerationrisk,crowdbehaviorrecognitionplaysanimportantroleinvisualsurveillanceandhasbeenhighlyconcernedbyrelevantauthorities.TheframeworkforcrowdbehaviorrecognitionisshowninFig.1.Thedynamicsofac
8、rowdischaracterizedandinfluencedbyanumberoffactors.Itisnotnecessaryt
