| 期刊名称： |
IEEE Transactions on Intelligent Transportation Systems |
| 全部作者： |
Baoding Zhou*,Qingquan Li,Qingzhou Mao,Wei Tu,Xing Zhang,Long Chen |
| 出版年份： |
2015 |
| 卷 号： |
16 |
| 期 号： |
5 |
| 页 码： |
2774 - 2785 |
| 查看全本： |
|
Indoormapsareintegraltopedestriannavigationsystems,anessentialelementofintelligenttransportationsystems(ITS).Inthispaper,weproposeALIMC,i.e.,ActivityLandmark-basedIndoorMappingsystemviaCrowdsourcing.ALIMCcanautomaticallyconstructindoormapsforanonymousbuildingswithoutanypriorknowledgeusingcrowdsourcingdatacollectedbysmartphones.ALIMCabstractstheindoormapusingalink–nodemodelinwhichthepathwaysarethelinksandtheintersectionsofthepathwaysarethenodes,suchascorners,eleva-tors,andstairs.Whenpassingthroughthenodes,pedestriansdothecorrespondingactivities,whicharedetectedbysmartphones.Afteractivitydetection,ALIMCextractstheactivitylandmarksfromthecrowdsourcingdataandclusterstheactivitylandmarksintodifferentclusters,eachofwhichistreatedasanodeoftheindoormap.ALIMCthenestimatestherelativedistancesbetweenallthenodesandobtainsadistancematrix.Basedonthedistancematrix,ALIMCgeneratesarelativeindoormapusingthemultidimensionalscalingtechnique.Finally,ALIMCconvertstherelativeindoormapintoanabsoluteonebasedonseveralreferencepoints.ToevaluateALIMC,weimplementALIMCinanofficebuilding.Experimentresultsshowthatthe80thpercentileerrorofthemappingaccuracyisabout0.8–1.5m.
