面向轨迹起止特征点数据的多比例尺可视化方法

发布时间：2018-06-13 13:00

本文选题：轨迹数据 + 多比例尺表达　；参考：《地球信息科学学报》2017年08期

【摘要】：本研究以北京市出租车GPS轨迹数据为例,建立了一种面向轨迹起止特征点(Origin-Destination,OD)的多比例尺可视化表达方法。首先,依据轨迹点描述信息提取OD特征点,并进行无效点清理与排除;然后,利用分布密度指标和辅助行政区划数据实施聚类分析,对OD数据分布空间进行区域划分;最后,定义参量统计各区域间OD数据隐含的流向特征,并设计专门符号进行可视化。其中,通过调整最小区域面积控制参数建立与街区、商圈、城区等不同层次地理单元相对应的区域划分,从而获得涵盖3种不同级别的OD数据多比例尺表达结果。试验结果表明,本文提出的方法能够对轨迹OD数据进行有效降维,获取不同尺度下区域间的车辆移动关系,对揭示车流人流时空交互模式及辅助决策有参考意义。
[Abstract]:Taking the GPS trajectory data of taxi in Beijing as an example, a multi-scale visual representation method for origin-Destation ODG (locus origin-Destation ODG) is established. Firstly, OD feature points are extracted according to the description information of track points, and the invalid points are cleaned up and excluded. Then, the distribution space of OD data is divided by cluster analysis using distribution density index and auxiliary administrative division data. Finally, the distribution space of OD data is divided. Parameters are defined to calculate the implied flow characteristics of OD data between regions, and special symbols are designed to visualize them. By adjusting the control parameters of the minimum area, the multi-scale representation results of OD data covering three different levels are obtained by establishing the regional division corresponding to geographical units at different levels, such as blocks, commercial circles, and urban areas. The experimental results show that the proposed method can effectively reduce the dimension of trajectory OD data and obtain the relationship of vehicle movement between different scales. It has reference significance for revealing the space-time interaction mode of traffic flow and auxiliary decision.
【作者单位】：信息工程大学;西安测绘研究所;北京大学遥感与地理信息系统研究所;国土资源部城市土地资源监测与仿真重点实验室;武汉大学资源与环境科学学院;
【基金】：国家自然科学基金项目(41401447) 国土资源部城市土地资源监测与仿真重点实验室开放基金项目(KF-2016-02-020) 国家高技术研究发展计划(“863”)项目(2015AA124103)
【分类号】：P208;P228.4

【相似文献】