中文摘要：

为了获取高精度的 DEM数据，采用了一种基于图像分割技术的方法对机载 LiDAR数据进行滤波处理。该方法针对原始激光点云数据，首先运用边缘检测技术进行点云分割，准确提取建筑物的边缘信息，确定地面种子点；然后按照一定的准则进行区域生长，最终得到裸露地面点集。针对几组不同地貌特征的城区实测数据进行滤波实验，通过定性和定量分析，验证了该算法不但计算速度快、精度高，而且还具有较强的稳健性。

英文摘要：

For the purpose of obtaining high precision DEM, this paper proposes a filtering algorithm of LiDAR data based on segmentation. Firstly, the data are divided into several segments and reliable seeds are selected according to edge detection, and then the ground is obtained by a certain growing rule. In order to test the filtering algorithms, the authors made experiments on several urban areas with different characteristics, and then analyzed the results qualitatively and quantitatively. The results show that the filtering method can effectively extract ground points from point cloud with high stability.