基于多密度聚类算法的路网提取

Road Network Extraction Based on Multi-density Clustering Algorithm

随着GPS等定位系统的迅速发展,使得路网提取有了新的发展方向,然而目前利用浮动车GPS轨迹数据提取路网的方法层出不穷,但是提取效果却并不能尽如人意。本文将基于网格密度因子的多密度聚类算法引入路网提取,通过该方法对疑似特征点进行聚类来提取道路特征点,来构建道路几何网络,并通过对比分析取得了不错的实验结果。

With the rapid development of G P S and other positioning systems, the road network has a new direction of development. However, the current floating vehicle G P S track data extraction network approach is endless, but the extraction effect is not satisfactory . In this paper, a multi-density clustering algorithm based on grid density factor is introduced into road network extraction. The road feature points are extracted by clustering the suspected feature points to construct the road geometry network, and the experimental results are obtained through comparative analysis.