众源GPS轨迹归并算法研究

发布时间：2019-01-10 08:50

【摘要】：随着移动互联网与移动终端的迅猛发展与普及,智能手机价格的降低,促使智能手机在大部分地区及人群中普及,并成为人们生活中不可或缺的工具。通过移动手机获取GPS(Global Positioning System,全球定位系统)轨迹数据变得越来越方便,规模也越来越庞大,因而出现了海量移动对象轨迹数据。而这些轨迹数据,产生了各种基于位置服务(Location Based Services,LBS)的应用。因此,如何从海量的GPS轨迹数据中自动提取、构建路网并更新路网信息,已成为当前研究的热点,该问题的解决在改善城市交通环境、车辆导航、景区旅游和灾害救助等方面扮演着越来越重要的角色。采用计算机软件、GIS(Geographic Information System或Geo-Information System,地理信息系统)等相关技术,考虑空间数据结构中矢量数据结构和栅格数据结构的特点,论文提出了一种众源GPS轨迹归并算法,该算法包括矢量数据栅格化、栅格数据二值化、栅格数据矢量化和轨迹中心线提取四个步骤。具体来讲,首先基于ArcGIS Engine组件,并采用C#语言实现了轨迹归并算法,然后在此基础上进行了4组实验,并利用Google Earth和ArcGIS Desktop软件对四组实验数据进行了可视化分析。通过以上的分析,论文得出了以下几点结论:(1)在轨迹数据预处理中,道格拉斯普克-线性简化(Douglas-Peucker,DP)算法对漂移点去除效果较好;具有噪音的基于密度的聚类(Density-Based Spatial Clustering of Applications with Noise,DBSCAN)算法对噪音点去除取得较好效果。(2)使用优化的矢量数据栅格化方法,将矢量数据转换成栅格数据可以取得较好效果;在输出归并后的轨迹中心线过程中,高斯滤波算法对轨迹进行平滑处理可以获取较好效果。(3)论文提出的众源GPS轨迹归并算法,能够准确提取轨迹路网并生成中心线。
[Abstract]:With the rapid development and popularization of mobile Internet and mobile terminals, the reduction of smartphone prices has promoted the popularity of smart phones in most areas and people, and has become an indispensable tool in people's lives. The acquisition of GPS (Global Positioning System, global positioning system (GPS (Global Positioning System,) trajectory data by mobile phone becomes more and more convenient, and the scale is becoming larger and larger, so a large number of mobile object trajectory data appear. These locus data, however, produce a variety of location-based service (Location Based Services,LBS) applications. Therefore, how to automatically extract, construct and update the road network information from massive GPS track data has become a hot topic of current research. The solution of this problem lies in improving the urban traffic environment and vehicle navigation. Scenic area tourism and disaster relief and other aspects play a more and more important role. By using computer software, GIS (Geographic Information System or Geo-Information System, Geographic Information system (GIS) and considering the characteristics of vector data structure and raster data structure in spatial data structure, a multisource GPS locus merging algorithm is proposed in this paper. The algorithm includes four steps: vector data rasterization, raster data binarization, raster data vectorization and locus centerline extraction. Specifically, the track merging algorithm is implemented based on ArcGIS Engine component and C # language, then four groups of experiments are carried out, and four groups of experimental data are visualized and analyzed by Google Earth and ArcGIS Desktop software. Through the above analysis, the paper draws the following conclusions: (1) in the trajectory data preprocessing, Douglas Puck linear simplification (Douglas-Peucker,DP) algorithm is better for the removal of drift points; The noisy density-based clustering (Density-Based Spatial Clustering of Applications with Noise,DBSCAN) algorithm has good effect on noise point removal. (2) an optimized vector data rasterization method is used. Good results can be obtained by converting vector data into raster data. In the process of output merging, Gao Si filter algorithm can get good results by smoothing the trajectory. (3) the multi-source GPS trajectory merging algorithm proposed in this paper can accurately extract the track network and generate the central line.
【学位授予单位】：成都理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P228.4

【参考文献】