基于浮动车轨迹数据的路网快速提取  

Rapid Extraction of Road Network Based on Floating Vehicle Trajectory Data

作　　者：米春蕾 彭玲 姚晓婧 陈六嘉 池天河 MI Chun-lei;PENG Ling;YAO Xiao-jing;CHEN Liu-jia;CHI Tian-he(Institute of Remote Sensing and Digital Earth,CAS,Beijing 100094;University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院遥感与数字地球研究所,北京100094 [2]中国科学院大学,北京100049

出　　处：《地理与地理信息科学》2019年第1期12-19,共8页Geography and Geo-Information Science

基　　金：国家科技支撑计划项目(2015BAJ02B00);国家自然科学基金项目(41701438)

摘　　要：浮动车轨迹数据包含丰富的路网信息,随着浮动车轨迹数据的逐渐公开,从中提取路网信息已成为可能。目前,大多数算法提取路网时,使用统一的阈值忽略了轨迹数据的密度差异,且只考虑了轨迹的形态没有考虑轨迹的方向,严重影响了提取结果的几何精确度和拓扑正确度。为此,该文提出了一种自适应半径质心漂移聚类方法,能根据轨迹密度、道路宽度自动调整聚类参数和利用轨迹方向实现道路拓扑连接。首先,通过自适应半径质心漂移聚类方法计算路网骨架点,采用小波聚类算法获取路网骨架点的方向集;然后,根据聚类半径和方向对骨架点进行递归连接,生成路网数据。利用深圳市福田区一天的浮动车轨迹数据进行了算法实验验证,将实验结果与栅格化方法、约束三角网方法的结果进行了定性定量评价分析。实验结果表明,该文算法提取的路网数据在几何精确度及拓扑正确度上都有明显的提高,且算法适合大数据处理。Floating vehicle trajectory data contains abundant road network information.With the gradual opening of floating vehicle trajectory data,it is possible to extract road network information from it.Currently,most road network extraction algorithms use unified thresholds to ignore the density difference of trajectory data,and only consider the trajectory shape without considering the direction of the trajectory,which seriously affects the geometric precision and topological accuracy of their results.Therefore,an adaptive radius centroid drift clustering method is proposed in this paper,which can automatically adjust clustering parameters according to the track density and the road width,using trajectory direction to complete the topological connection of roads.Firstly,the skeleton points of road network are calculated by adaptive radius centroid drift clustering method,and the direction sets of skeleton points are obtained by wavecluster algorithm.Then,the skeleton points are recursively connected based on clustering radius and direction to generate road network data.The algorithm is verified by the floating vehicle trajectory data of a day in Futian District,Shenzhen.The experimental results are qualitatively and quantitatively analyzed with ones of the raster method and the constrained triangulation method.It indicates that the road network data extracted by this algorithm has a significant improvement in geometric precision and topological accuracy,and which is suitable for large data processing.

关 键 词：浮动车轨迹数据 路网提取 自适应半径质心漂移聚类 小波聚类 

分 类 号：P228.4[天文地球—大地测量学与测量工程] U491.13[天文地球—测绘科学与技术]