基于步行轨迹的复杂道路中心线提取方法

传统的基于GPS轨迹的路网提取多关注使用车载GPS轨迹数据提取城市车道级路网，忽略了校园、社区、景区等小范围区域内的复杂道路信息，该文提出了一种基于步行GPS轨迹数据的复杂道路中心线提取方法.在数据清洗的基础上，基于轨迹点分布密度进行栅格化，然后采用Zhang-Suen快速细化算法对栅格路网细化得到道路中心线，最后采用改进的追踪法实现路网矢量化.以华中师范大学校园为试验区域，提取出了人车混行、小路交错的校园路网，证明了该方法的有效性.

Method to extract the center line of a complex road based on walking trajectory

Traditional road network extraction based on GPS trajectory focuses more on using vehicle-mounted GPS trajectory data to extract urban lane-level road network， while ignoring complex road information in small areas such as campus， community and scenic spots. Therefore， this paper proposed a method to extract complex road center line based on walking GPS trajectory data. Firstly the data was cleaned， then GPS trajectories were rasterized based on the distribution density of points， and zhang-suen rapid refinement algorithm was used to refine the grid network to obtain the road center line. Finally， the improved tracking method was used to realize the vectorization of road network. The paper took the campus of Central China Normal University as the experimental area， and extracted the campus road network with mixed vehicles and small roads， which proves the effectiveness of the method.