蚁群算法的建筑立面点云数据提取

Ant Colony Algorithm for Building Facade Points Extraction

摘要: 提出了一种基于蚁群算法的地面激光扫描数据建筑物立面提取方法,该方法可以有效地区分沿街LiDAR扫描景观数据中建筑物立面和位于其前方的树木、街灯、行人、停靠车辆等遮挡物。三组真实的地面激光点云的实验结果表明,该方法能准确、有效地提取建筑物立面点云数据。

Abstract: We present an automatic building facade extraction approach based on terrestrial laser scanning points,which effectively use the ant colony algorithm to remove the occlusions and get rid of the error points caused by windows.The method is tested with terrestrial point clouds of three different street scenes.The results show that the method is effective.