基于轨迹方向特征的道路交叉口提取算法

Road Intersection Extraction Algorithm Based on Trajectory Direction Features

作为城市路网骨架的定位节点,道路交叉口在城市交通地理信息服务中扮演着重要角色。浮动车GPS轨迹数据包含丰富的路网信息,具有成本低、时效性高、获取方便等优点,已被广泛应用于道路交叉口提取的相关研究中。以浮动车GPS轨迹数据为研究对象,提出了一种道路交叉口提取算法;同时考虑交叉口区域浮动车行驶方向变化和行驶方向的差异性,分别从单条轨迹和全部轨迹出发提取交叉口附近的转向点集,并对这两组转向点集进行聚类和整合,得到道路交叉口位置信息。该算法在公开的浮动车轨迹数据集上进行了测试,结果表明算法能有效识别道路交叉口。

As the positioning nodes of urban road network skeleton,road intersection plays an important role in urban traffic geographical infor-mation service.The GPS trajectory data of floating car contains rich road network information,has the advantages of low cost,high timeliness and convenient acquisition,and has been widely used in the road intersection extraction research.In this paper,we proposed a road intersection extraction algorithm based on GPS trajectory data of floating cars.Considering the change of driving direction and the difference of driving direc-tion of floating cars in the intersection area at the same time,we extracted the set of turn-points near the intersections from a single trajectory and all trajectories respectively,and obtained the location information of road intersections by clustering and integrating these two sets of turn-points.Finally,we tested the algorithm on a publicly available floating car trajectory data set.The result shows that the algorithm can effectively identify road intersection.