Target Tracking with Target State Dependent Detection

Target Tracking with Target State Dependent Detection

ID:39772715

大小:421.88 KB

页数:12页

时间:2019-07-11

Target Tracking with Target State Dependent Detection_第1页
Target Tracking with Target State Dependent Detection_第2页
Target Tracking with Target State Dependent Detection_第3页
Target Tracking with Target State Dependent Detection_第4页
Target Tracking with Target State Dependent Detection_第5页
资源描述:

《Target Tracking with Target State Dependent Detection》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、IEEETRANSACTIONSONSIGNALPROCESSING,VOL.59,NO.3,MARCH20111063TargetTrackingWithTargetStateDependentDetectionTaekLyulSong,Member,IEEE,DarkoMusickiˇ,Member,IEEE,andKimDaSolAbstract—Targettrackingalgorithmsusuallytreattheprob-measurementperscan.Thenumberofpossiblemeasurementse-abilityofdet

2、ectionasindependentofthetargetstate.Inmostquencesgrowsexponentiallyintimewhileeachsequencewillcases,thisassumptionisnottrue,withsubsequentdegradationresultinatargettrajectorystateestimate.Thus,varioustech-inthetargettrackingperformancefrombothexpectedandop-niquesareusedtocontroltheirnu

3、mber[1].Onewayofdealingtimallevels.OnetypicalexampleistheDopplerfrequencybasedclutterrejection,theotherisobfuscation(shadowing)ofgroundwithcomplexityistoreducetheprobabilitydensityfunctionbasedtargets,andthethirdisantijammingnotchfiltering.This(pdf)ofeachtracktoasingleGaussian;theseappr

4、oachesaredependenceoftheprobabilityoftargetdetectiononthetargettermedsingle-scantrackersandincludetrackersbasedonproba-trajectorystatemodulatesthemeasurementlikelihood,which,bilisticdataassociation(PDA)[2],[3],andprobabilisticnearestinturn,introducesmeasurementnonlinearity.Inthispaper,

5、weneighborfiltering[4],[5].Multiscantrackersretainestimatesfirstpresentageneralalgorithmfortargettrackinginclutterwhentheprobabilityofdetectionistargetstatedependent,andbasedonmultiplemeasurementhistoriesandtheirpdfsareap-thenproceedtoanalgorithmwherebothtargetstateestimateandproximatedb

6、yGaussianmixtures(GMs);examplesincludevar-theprobabilityofdetectionaremodeledbyGaussianmixtures.iousformsofmultiplehypothesistracker(MHT)[6],[7],andTheprobabilityoftargetexistenceisrecursivelyupdatedasthetracksplitfilters[8].Possiblepresenceofmultipletargetsaddstrackqualitymeasureusedfo

7、rfalsetrackdiscrimination.Aanotherlevelofcomplexity.Forreasonsofclarityandbrevity,two-sensor-basedgroundmaneuveringtargettrackingincluttersimulationvalidatesthisapproach.weconcentrateinthispaperonsingletargettracking,withtheexceptionofSectionIV-AwithabriefoverviewofapossibleIndexTerm

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。