基于自适应轨迹聚类的城市路网提取与更新方法  

Extraction and updating method for urban road network based on adaptive trajectory clustering

作　　者：彭程[1] 唐建波[1,2] 彭举 梅小明 陈雪莹[1] 姚志鹏 PENG Cheng;TANG Jianbo;PENG Ju;MEI Xiaoming;CHEN Xueying;YAO Zhipeng(School of Geosciences and Info-physics,Central South University,Changsha 410083,China;Hunan Geospatial Information Engineering and Technology Research Center,Changsha 410083,China)

机构地区：[1]中南大学地球科学与信息物理学院,长沙410083 [2]湖南省地理空间信息工程技术研究中心,长沙410083

出　　处：《时空信息学报》2023年第2期209-217,共9页JOURNAL OF SPATIO-TEMPORAL INFORMATION

基　　金：国家自然科学基金项目(42271462,42171441);湖南省自然科学基金项目(2021JJ40727,2022JJ30703,2020JJ4749);中南大学中央高校基本科研业务费专项资金资助项目(2022zzts105);湖南省研究生科研创新项目(CX20220193)。

摘　　要：由于城市路网更新频繁及传统地图更新方法的局限性,很多城市路网地图数据无法及时更新。基于移动轨迹数据的路网地图更新是实现城市路网地图快速、高效更新的有效途径。虽然目前基于移动轨迹数据发展了一些基于栅格化的方法、基于聚类的方法和增量化的方法来提取与更新城市路网,但这些方法对于参数依赖性较强,且难以准确提取较为复杂的道路形状。因此,本文结合移动轨迹数据的特点及道路网的结构特征,提出了一种基于移动轨迹匹配的城市路网地图动态更新方法。首先,通过轨迹点的位置和方向属性进行轨迹点–路段匹配;然后,对未匹配的轨迹点采用顾及方向约束的空间自适应聚类算法进行聚类,并对每个聚类进行最优曲线拟合,获得新增道路;最后,以城市浮动车轨迹数据为例进行实验与对比分析。实验结果表明,与现有方法相比,本文方法需要的参数较少,且能够处理复杂形状道路的提取,具有较高的运行效率。Due to the frequent updates of urban roads infrastructure and the limitations of traditional road updating methods,many urban roads cannot be updated promptly.To date,road network updating based on moving trajectories has proven to be an effective way to realize the rapid and efficient updating of urban roads.While some methods based on GPS trajectories,such as grid-based methods,clustering-based methods and incremental methods,have been developed,these methods tend to be sensitive to parameters.Moreover,it is still difficult for these methods to accurately extract the incremental roads with complex structures.In light of these concerns,we propose an automatic urban road updating method based on moving trajectory matching.This method takes into account both the characteristics of the moving trajectory data and the structural characteristics of the urban roads.In this paper,the location and direction of the trajectory points are firstly used to make point-to-line matching.Then,an adaptive spatial clustering algorithm,which incorporates the direction constraint,is developed to segment the unmatched points into tracks,and the optimal curve fitting is performed for each cluster to obtain the center lines of the incremental roads.Finally,the urban floating car trajectory data was used to test the proposed method.When compared with commonly used rasterization methods and center track point fitting methods,our proposed method can better identify new additive roads in the urban road network.This paper’s primary contributions are as follows:①We focus on extracting localized changes of the existing road network using a partial trajectory map matching strategy.This avoids the reconstruction of the entire urban road network,thereby enhancing the efficiency of real-time update of the urban road network;②Based on the spatial proximity and direction constraints,the local matching method of massive track-road network is proposed.This approach offers high matching accuracy and lower computing complexity;③The adaptive spat

关 键 词：轨迹数据 地图匹配 自适应聚类 城市路网 路网更新 

分 类 号：P208[天文地球—地图制图学与地理信息工程] U495[天文地球—测绘科学与技术]