资源描述：
《一种基于背景模型的运动目标检测与跟踪算法》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。
1、Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/246875977MovingobjectdetectionandtrackingbasedonbackgroundsubtractionArticleinProceedingsofSPIE-TheInternationalSocietyforOpticalEngineering·September2001
2、DOI:10.1117/12.441618CITATIONSREADS378833authors,including:HaizhouAiGuangyouXuTsinghuaUniversityTsinghuaUniversity131PUBLICATIONS3,178CITATIONS253PUBLICATIONS2,293CITATIONSSEEPROFILESEEPROFILESomeoftheauthorsofthispublicationarealsoworkingontheserelatedp
3、rojects:863-306-ZT03-01-1AStudyonRobustFaceRecongnition&VerificationViewprojectAllcontentfollowingthispagewasuploadedbyGuangyouXuon15October2015.Theuserhasrequestedenhancementofthedownloadedfile.第31卷第4期信息与控制Vol.31,No.42002年8月InformationandControlAug.,200
4、2文章编号:1002-0411(2002)04-315-05一种基于背景模型的运动目标检测与跟踪算法刘亚艾海舟徐光佑(清华大学计算机科学与技术系北京100084)摘要:本文提出了一种静止摄像机条件下的运动目标检测与跟踪算法.它以一种改进的自适应混合高斯模型为背景更新方法,用连通区检测算法分割出前景目标,以Kalman滤波为运动模型实现对运动目标的连续跟踪.在目标跟踪时,该算法针对目标遮挡引起的各种可能情况进行了分析,引入了对运动目标的可靠性度量,增强了目标跟踪的稳定性和可靠性.在对多个室外视频序列的实验
5、中,该算法显示了良好的性能,说明它对于各种外部因素的影响,如光照变化、阴影、目标遮挡等,具有很强的适应能力.关键词:背景模型;混合高斯模型;Kalman滤波;运动目标检测与跟踪中图分类号:TP13文献标识码:BMOVINGOBJECTDETECTIONANDTRACKINGBASEDONBACKGROUNDSUBTRACTIONLIUYaAIHai-zhouXUGuang-you(DepartmentofComputerScienceandTechnology,TsinghuaUniversity)Ab
6、stract:Anapproachtodetectingandtrackingmovingobjectswithastaticcameraispresentedinthispaper.AmodifiedmixtureGaussianmodelisusedastheadaptivebackgroundupdatingmethod.Foregroundobjectsaresegmentedbasedonanimprovedbinaryconnectedcomponentanalysis.Kalmanfilt
7、eringisusedforobjecttracking.Todealwiththeproblemsofocclusionbetweenobjectsintracking,varioussituationsareanalyzedandameasureofreliabilityofmovingobjectsisadoptedwhichmakesthetrackermoreeffective.Experimentsonseveraloutdoorvideostreamsthatshowconvictiveo
8、bjectdetectionandtrackingperformancedemonstrateitsstrongadaptabilitytolightingchanges,shadowsandocclusions.Keywords:backgroundmodeling,mixtureGaussianmodel,Kalmanfiltering,movingobjectdetectionandtracking[2]1引言(Introductio
