车载联合机载LiDAR点云数据的建筑物立面精细分割

从数据量庞大且散乱的车载LiDAR点云中分割出建筑物立面数据是一项繁琐而艰巨的工作。本文提出一种结合机载LiDAR点云的车载LiDAR点云建筑物立面分割方法。该方法在空-地点云严格配准的基础上,从机载LiDAR点云中分割出每栋建筑物的顶部点云,提取建筑物顶部外轮廓线并进行规则矢量化处理,设置轮廓线缓冲区实现立面点云的粗分割;再采用基于稳健特征值的平面拟合法对单栋建筑物的每个立面进行去噪滤波,实现建筑物立面的精细分割。试验结果证明了该算法对城市场景中车载LiDAR点云处理的有效性。

Fine Segmentation of Building Facade Combined Mobile LiDAR with Airborne LiDAR Point Cloud Data

The segmentation of the building facade point data from huge and scattered building point cloud is a complicated and difficult task.In this paper,a method of mobile LiDAR point cloud building facade segmentation combined with airborne LiDAR point cloud is proposed.Firstly,on the basis of the strict registration of the airborne LiDAR point cloud and mobile LiDAR point cloud,the roof point cloud of each building is segmented from airborne LiDAR point cloud.The outline of each roof is extracted and then regularized and vectorized.Then the coarse segmentation of the facade point cloud is achieved by setting contour line buffer.Finally each facade of the single building is filtered and denoised by using the method of robust eigenvalue plane fitting,which can achieve fine segmentation of the building facade.Experimental results demonstrate the effectiveness of the proposed algorithm for urban scene mobile LiDAR point cloud processing.