基于LIDAR数据的道路平面线形拟合方法研究

阐述了LIDAR数据滤波与分类方法,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,结合LIDAR数据的特点提出了一种针对带状离散点云的定步长径向搜索方法以提取道路中轴点列并对点列进行道路平面线形拟合与恢复,试验结果表明,基于既有知识数学形态学分类方法能较好地从LIDAR地面数据点云中提取道路信息;从机载LIDAR数据的离散带状点云中拟合公路平面线形,所采用的搜索算法简单快速,运算效率高.

Research on Fitting Method of Road Horizontal Alignment Based on LIDAR Data

The filtering and classification methods based on LIDAR data were intreduced. Based on the existing knowledge of mathematical morphology classification, a new method of concentrated extracting road information at original discrete points was put forward. According to the characteristic of the LIDAR data, a radial search method with fixed length of stride for belt-shaped scattered cloud was proposed to extract the range of points on road centerline for fiting and restoring the road horizontal alignment. The results of experiment show that (1) the method based on the existing knowledge of mathematical morphology classification can extract information of road from clouds of LIDAR data; (2) the searching algorithm is simple and effective for fitting the road horizontal alignment from the beh-shaped scattered cloud of on-board LIDAR data.