Vehicle-based　crowdsourcing　(VBC)　is　a　sourcing　model　in　which　crowdsourcers　hire　a　bunch　of　vehicle　users　to　accomplish　sensing　tasks.　The　design　of　the　incentive　mechanism　in　VBC　is　necessary　because　the　amount　of　the　sensing　tasks　accomplished　depend　on　the　scale　and　the　willingness　of　the　participants.　In　general,　every　sensing　task　has　two　dimensions,　which　are　the　time　dimension　and　place　dimension.　That　is　to　say,　each　task　in　VBC　corresponds　to　different　Places　of　Interest(PoIs)　and　time　slot.　However,　most　of　the　existing　incentive　mechanisms　only　allow　users　to　enter　the　system　one　by　one　without　considering　both　the　two　characteristics　of　the　sensing　task,　only　take　into　account　the　user＇s　margin　contribution　value　rather　than　the　margin　contribution　ratio.　On　the　contrary,　we　fully　consider　the　user＇s　margin　contribution　ratio,　which　takes　into　account　the　system＇s　requirements　for　the　task　and　the　ability　of　the　user　to　do　the　task.　In　addition,　our　mechanism　can　be　effectively　applied　to　a　time-sensitive　task　scenario.　In　this　paper,　we　propose　an　online　incentive　mechanism,　RIM,　with　the　properties:　truthfulness,　budget　feasible,　computational　efficiency,　individual　rationality,　consumer　sovereignty.　With　extensive　simulation,　we　show　that　the　results　and　the　theoretic　analysis　are　in　accord.　In　addition,　the　proposed　mechanism　outperforms　the　existing　methods.　©　2018　IEEE.