融合OSM和车载激光扫描数据的建筑物三维快速提取

3D Buildings Fast Detection by Fusion of Open Street Map Data and MLS Point Clouds

建筑物作为一种重要的地理空间要素,在城市规划与建设、居住环境、交通管理、房地产以及灾害损失评估等领域占有极为重要的地位。车载激光扫描(Mobile Laser Scanning, MLS)数据是建筑物三维模型重建的主要数据源之一,本文针对基于MLS数据建筑物的提取耗时过长和错分漏提率高等问题,提出了一种融合Open Street Map (OSM)和MLS点云数据的三维建筑物快速提取方法。该方法包含数据预处理、改进的ICP精配准、OSM与MLS数据的信息融合及辅助提取四个步骤,实验结果表明该方法可以有效提升效率,处理时间可缩短50%以上;同时该方法还可有效避免误提取问题,对漏提取问题也有所改善,因此提升了建筑物提取的精度。两种数据的融合,实现了双方的优势互补,为建立位置精准、关系正确、几何细节和属性信息丰富,方便快速更新的城市三维模型,提供了数据支撑。

As a geospatial element, building plays an extremely important role in fields of urban planning and construction, residential environment, traffic management, real estate industry, and hazard assessment. Mobile Laser Scanning (MLS) data is one of the main data sources for 3D-models reconstruction of buildings. This paper proposes a fast method of 3D-building extraction from MLS point clouds, by fused with Open Street Map (OSM) data, to solve the problems of time-consuming and building mistaken extraction. The method includes four steps: data preprocessing, improved ICP fine registration, information fusion of OSM and MLS, and assisted building extraction. The experimental results show that this method can effectively improve the efficiency, the processing time is shortened by more than 50%, and it is effective to avoid the problem of building mistaken and missing, therefore improves the precision and accuracy of building extraction. The fusion of OSM and MLS data can provide data support for 3D urban models with accurate location, correct relations, geometric details, and rich semantic information.