论文部分内容阅读
机载激光雷达(LIDAR)技术的出现为道路特征的获取提供了新的途径。在分析现有道路提取现状的基础上,针对激光雷达数据的特点以及单一尺度下道路中线提取方法的不足,提出一种基于多尺度追踪的道路中线提取方法。该方法首先采用逐步约束的方法进行道路激光点的提取,包括高程约束、强度约束以及区域点密度和区域面积的约束等;然后基于道路点云生成的不同尺度距离影像的形态学细化结果,采用多尺度追踪的方法实现道路中线的提取,其中多尺度追踪方法由大尺度道路中线的迭代追踪以及小尺度道路中线的启发式追踪两部分组成。最后采用实地数据进行验证,结果表明:该方法能有效地从LIDAR点云中提取道路中线信息,并具有较好的精度。
The emergence of LIDAR technology has provided a new way for the acquisition of road features. Based on the analysis of the present situation of the existing road extraction, aiming at the characteristics of laser radar data and the lack of the method of extracting the road center line under a single scale, a method of road center line extraction based on multi-scale tracking is proposed. Firstly, the method of step-by-step constraint is used to extract the laser points on the road, including the elevation constraints, the intensity constraints and the constraints of the area density and area. Then based on the morphological refinement results of the images at different scales generated by the road point cloud, Multi-scale tracking method is adopted to extract the centerline of the road. The multi-scale tracking method consists of two parts: the iterative tracking of large-scale road centerline and the heuristic tracking of small-scale road centerline. Finally, the field data is used to verify the results. The results show that this method can effectively extract the road centerline information from LIDAR point cloud and has better accuracy.