基于车载激光扫描数据的窗户提取与重建技术

The Study on Extracting and Rebuilding Windows from Vehicle-borne Laser Scanning Data

借鉴投影点密度的思想,提出了一套从车载激光扫描数据中提取和重建窗户的方案。首先,将建筑物立面点云投影到立面平面,通过建立不规则三角网(TIN),比较三角形边长来提取窗户边缘点。然后根据边缘点的分布建立格网,计算格网内的投影点密度,分类出代表窗户区域的子块并对同一窗户内的子块进行聚类。将得到的窗户在立面平面上的二维坐标代入立面平面方程,并结合窗户的厚度进行解算得到窗户立体模型各角点的三维坐标。最后利用方案对建筑物立面点云数据进行了具体处理并给出了窗户的提取和三维重建的效果。

A schema of extracting and rebuilding windows from point cloud data of Vehicle-borne Laser Scanning System（VLSS） with the use of projection density was brought out. Firstly, the building facade point clouds were projected to facade plain ,then Triangulated Irregular Network（TIN） was erected and window edge points were extracted by comparing side length of triangles. Afterwards, grids were erected through the distributing of edges points, and the sub blocks representing windows were classified and clustered after the projection density of grids were accounted. In the following,coordinates of window corner points on the facade plain were introduced to faqade plain function to compute the 3-D coordinates of corner points with the help of thickness of windows. At last, building facade point clouds are processed using the schema in this paper and extracting and 3-D rebuilding effects are presented.