This　paper　presents　a　matching　method　for　rapidly　acquiring　corresponding　points　in　the　hybrid　vision　system.　Firstly,　an　elementary　introduction　to　the　research　of　the　hybrid　vision　system　is　given.　Secondly,　the　Scale　Invariant　Feature　Transform　is　used　to　obtain　the　initial　matching　point　set.　Thirdly,　we　describe　the　hybrid　vision　system　model,　and　then　we　analyze　the　epipolar　geometry　of　this　system.　Based　on　this,　we　develop　a　novel　matching　method　to　eliminate　the　mismatched　points　and　enhance　the　accuracy　of　matching.　Finally,　experimental　results　prove　the　effectiveness　of　our　proposed　method.　©2012　IEEE.