利用路段分类识别复杂道路交叉口

发布时间：2018-06-29 15:28

本文选题：道路结构识别 + 道路交叉口　；参考：《武汉大学学报(信息科学版)》2016年09期

【摘要】：道路网数据中微观结构的识别对于多尺度路网建模、步行导航等至关重要。复杂道路交叉口是重要的道路微观结构之一,针对目前道路复杂交叉口基于几何形状描述与图形匹配识别方法存在的不足,从复杂交叉口识别与化简的角度出发,提出了一种利用路段分类进行复杂道路交叉口识别与化简的方法。该方法首先通过点密度聚类的方法对道路交叉口进行定位,然后利用路段的规模、形状和属性等特征构建特征空间,将交叉口的识别作为一种区分主干路段与辅助路段的两类分类问题,利用支持向量机的方法对交叉口区域内的路段进行分类,从而完成交叉口的识别与化简。利用开放街道地图(OpenStreetMap)数据进行实验,结果表明,该方法能够有效地识别道路交叉口。
[Abstract]:The identification of microstructure in road network data is very important for multi-scale road network modeling and walking navigation. Complex road intersection is one of the most important micro structures of roads. In view of the shortcomings of the current methods of road complex intersection recognition based on geometric shape description and graph matching, the recognition and simplification of complex intersection are considered. This paper presents a method of identifying and simplifying complex road intersections by using road section classification. The method firstly uses the point density clustering method to locate the road intersection, and then constructs the feature space by using the characteristics of the road section, such as size, shape and attribute, etc. The recognition of intersection is regarded as a kind of classification problem that distinguishes the main road section from the auxiliary road section, and the method of support vector machine is used to classify the road section in the intersection area, so as to complete the identification and simplification of the intersection. The experimental results show that the method can effectively identify the road intersection by using the OpenStreetMap data.
【作者单位】：信息工程大学;96633部队;
【基金】：国家863计划(2012AA12A404) 国家自然科学基金(41571399,41201391,41071297,41201469)~~
【分类号】：P208

【相似文献】