论文部分内容阅读
电子地图数据是各种智能交通系统的基础数据,对方便人们交通出行、解决城市交通拥堵问题具有重要的意义。其中,道路网络数据又是电子地图的重要组成部分。传统的道路网络数据更新方法往往需要耗费大量人力物力,因此,从其他数据源中如遥感影像数据、LIDAR数据等提取道路网络的相关研究,已成为国内外的研究重点。车载定位设备的广泛应用使车辆轨迹数据的大量获取成为可能,轨迹数据是对车辆行驶路径的完整记录,同时也是道路网络几何特征的直接反映。当轨迹数据量足够大时,则可利用其构建路网,用于更新或修正现有地图上的路网空间信息。针对车辆轨迹数据的特点和道路网络的特性,本文提出一种细化的道路网络几何特征提取方法。车辆轨迹数据是矢量数据,将其转换为栅格数据后,就可采用图像细化的方法处理。图像细化可以在保持原图像拓扑结构不变的情况下,快速地提取出图像的中心像元,并且有效去除冗余信息。本文以上海陆家嘴的车辆轨迹数据为例进行了实验,结果表明,利用车辆轨迹数据构建路网不仅可行,而且简单、高效,取得了良好的效果。
Electronic map data is the basic data of various intelligent transportation systems, which is of great significance to facilitate people’s traffic travel and solve the problem of urban traffic congestion. Among them, the road network data is an important part of the electronic map. Traditional data updating methods of road network often require a lot of manpower and resources. Therefore, the research on extracting road network from other data sources such as remote sensing image data and LIDAR data has become the research focus at home and abroad. The extensive application of vehicle positioning equipment makes it possible to obtain a large amount of vehicle trajectory data. The trajectory data is a complete record of the vehicle driving path and is also a direct reflection of the geometric characteristics of the road network. When the amount of trajectory data is large enough, it can be used to construct the road network, which is used to update or modify the road network space information on the existing map. According to the characteristics of vehicle trajectory data and the characteristics of road network, this paper presents a detailed method of extracting geometric features of road network. Vehicle trajectory data is vector data, convert it to raster data, you can use image refinement method to deal with. Image refinement can quickly extract the central pixel of the image while keeping the original image topology unchanged, and effectively remove the redundant information. This paper takes Shanghai Lujiazui vehicle trajectory data as an example to conduct an experiment. The results show that it is not only feasible, but also simple and efficient to construct the road network using the vehicle trajectory data, and achieved good results.