Abstract: Terrestrial laser scanner can quickly and precisely get all point clouds of a building, but it’s a time?consuming work to get key feature data which work for building maintenance and structure health assessment from these point clouds. In order to change now manual interaction work style, an efficient algorithm is presented which can quickly get key building profiles from huge point cloud, its basic ideas and workflow are described, and relative software tool is developed which can make building profiles with high productivity. Practical application shows that presented method has higher productivity and higher profile-making precision than existing building profile making methods.

Key words: point cloud, building profile, building maintenance, structure health assessment

摘要： 三维激光扫描技术可以快速而精确地获取既有建筑的完整点云数据，但要从点云中获得修缮设计及结构健康评估所需要的特征信息仍然是一个极为耗时耗力的工作。为改善目前纯手工交互的作业现状，提出了一种从点云中快速获取建筑特征剖面数据的算法，阐述了其基本思想与作业流程，并研发了基于该算法的具有较高自动化水平的剖面图制作软件工具。实际应用证明，软件在面对海量点云数据处理时，能够显著提高作业效率及提升成果制作质量。

关键词: 点云, 建筑剖面图, 建筑修缮, 结构健康评估