An　accurate　navigation　system　is　an　essential　and　important　part　for　the　mobile　robot.　The　recent　appearance　of　low　cost　RGBD　cameras　has　made　3D　point　clouds　together　with　RGB　information　easy　accessible,　and　they　have　been　widely　applied　in　many　applications.　Relative　poses　of　a　mobile　robot　can　be　estimated　from　consecutive　visual　information.　However,　such　incremental　registration　methods　still　suffer　from　accumulated　errors　which　makes　the　estimated　trajectory　as　weird　as　by　only　using　wheel　mounted　encoders.　In　contrast,　we　introduce　a　novel　and　inexpensive　sensor　fusion　based　approach　to　solve　the　robot　localization　problem.　The　key　idea　is　to　use　visual　gyroscope　as　a　complementary　source　for　robot　heading　estimation.　Aided　with　constraints,　the　unscented　Kalman　filter　is　used　for　robot　pose　estimation.　A　field　experiment　has　been　carried　out　in　order　to　verify　the　introduced　method.　Accordingly,　the　3D　map　of　the　environment　is　also　presented　based　on　the　estimated　robot　trajectory.　©　2014　IEEE.