利用既有知识渐近数学形态学提取LiDAR数据中道路信息方法研究

为了避免由于内插和构网或是局限于局部分析所带来的地形失真,本文根据道路先验知识的获取只需对实验道路进行统计学分析,无论从技术上,还是仪器设备上都简单、方便、快捷;在各种LIDAR数据滤波与分类方法的基础上,提出了一种在原始离散点集中提取道路信息的基于既有知识数学形态学分类方法。实验结果显示,为探索LIDAR数据在交通工程中的应用,加快道路的信息化建设,提供了一种可供借鉴的方法。

Method of extraction road information in LIDAR data based on the existing knowledge of asymptotic mathematical morphology

To avoid the terrain distortion caused by the interpolation and limited to local analysis or network configuration,and to ensure simple,convenient and fast operation no matter with technology or equipment and instruments by making a statistical analysis for the tested road to access a priori knowledge of the road,a new method about concentratively extracting road information among the original discrete points based on the existing knowledge of mathematical morphology classification was introduced,which was brought up from a variety methods of filtering and classification of LIDAR data.The results showed that it provided a way for reference in the exploration of using LIDAR data in traffic engineering and speeding up road information technology construction.