中文摘要：

针对道路面三维激光扫描的特点，结合点云法向量分布特征，提出了基于法向量模糊聚类的点云数据滤波算法。通过对3种不同类型实测道路面数据的实验表明，该算法对于路面扫描过程中产生的非路面冗余数据有较好的滤波效果，为点云数据滤波提供了一种新的手段。

英文摘要：

With the characteristic of 3D laser scanning on road and the distribution analysis of points cloud＇s normal vectors, the filtering algorithm based on fuzzy clustering of normal vectors is proposed. Three different types of real road data are put into experiment. It proves that the algorithm gives good filtering effect to non-road data produced in the road scanning. The algorithm provides a new method for filtering points cloud data.