Light　transition　model　in　homogenous　fog　has　extensive　applications.　Nowadays　it　is　used　to　estimate　scene　distance　and　used　for　3-D　reconstruction　especially　in　traffic　scene.　This　is　because　the　whole　road　could　be　identified　as　uniform.　However,　the　foundation　requite　for　these　application　is　the　current　weather　condition　is　the　homogenous　fog　weather　condition　which　is　presupposed　and　not　identified　by　most　of　the　authors.　A　methodology　for　homogenous　fog　weather　condition　recognition　for　transport　scene　based　on　texture　analysis　and　light　transition　model　is　presented　in　this　paper.　Firstly,　a　narrow　area　containing　the　sky　and　road　is　extracted　based　on　texture　analysis.　This　narrow　area　should　be　consistent　from　the　top　to　bottom　of　scene　background.　Secondly,　mean　value　for　each　line　is　calculated　and　saved.　We　call　these　data　Mean　Pixel　Values　(MPV).　These　values　are　sorted　from　the　top　to　the　bottom　and　fitted　by　mathematics　function.　Finally,　the　mathematic　function　style　is　compared　with　exponential　function.　If　the　mathematic　function　we　calculated　is　an　exponential　function,　then　the　current　weather　condition　is　homogenous　fog　condition,　otherwise,　other　weather　conditions.　We　test　our　algorithm　by　using　traffic　scene　in　four　weather　conditions.　Experimental　results　verify　the　effectiveness　of　the　algorithm.　©　2012　Huazhong　Univ　of　Sci　&　Tec.