Abstract: Two　types 　of 　approach　called 　re gion－growing　and　random　sample 　consensus　have　been　proposed 　for 　automatic　building　roof 　extraction．They　bothhoweverhave　drawbacks．In 　this　paperan hybridized　method　is 　proposed 　to 　take 　advantage　of 　both　al gorithmsstren gths 　so 　that 　building　roofscan　be　extracted　more　precisel y　and　efficientl y．Firstwe　calculate 　the　normal　and　rough　features from　LiDAR　data 　for 　the 　coarse 　extraction　of 　building　roofs．Second precise 　roof－extraction 　is 　performed　using　an　extended　RANSAC　method　which　takes 　the 　coarse 　extraction　results 　as 　the　priori knowledgeand 　inte grates 　a　re gion 　growing　method．Finallyapoll 　strate gy 　is 　adopted 　to 　solve 　the competition 　problem．The 　experimental 　results 　show　that 　our　method　can　extract 　intact 　building　roofs in 　a　hi ghl y　automated　manner．