Object　behavior　analysis　is　important　in　traffic　behavior　analysis.　In　this　paper,　we　propose　a　novel　approach　to　analyzing　object　motion　behavior.　First　we　build　the　background　probability　model,　then　background　subtraction,　binarization,　median　filtering　and　morphological　operation　are　implemented　in　sequence　to　obtain　the　candidate　regions.　Next,　the　sum　of　gradient　is　used　for　shadow　removal.　In　the　candidate　region　the　true　object　is　detect　and　their　behavior　is　analyzed.　The　experimental　results　confirm　the　efficiency　of　proposed　approach.　©　2012　IEEE.