We　propose　an　effective　localization　and　map　building　method　for　mobile　robot　using　a　monocular　camera　and　Laser　Range　Finder　(LRF).　To　compensate　for　the　insufficient　information　gathered　by　LRF,　we　improved　a　state　of　the　art　vision-based　SLAM　algorithm　so-called　parallel　tracking　and　mapping　(PTAM)　system　which　uses　an　optimized　parallel　implementation　for　localization　of　mobile　robot　with　high　accuracy,　and　integrate　the　laser　scanning　data　into　local　map　using　an　effective　combination　approach.　A　reliable　spatial　model　is　produced　for　mobile　robot　path　planning　or　obstacle　avoidance.　Compared　with　the　traditional　method,　the　proposed　approach　can　acquire　more　features　in　environment　of　mobile　robot　and　build　the　map　with　high　speed　and　improved　accuracy　in　localization.　Finally,　the　effectiveness　of　proposal　is　verified　by　some　real　experimental　results.　©　2011　IEEE.