道路网匹配过程往往采用全局遍历的搜索模式,这种模式会影响匹配效率。针对这一问题,提出一种利用道路分类的层次迭代匹配新方法。首先,依据拓扑关系对道路进行分类,并按道路类型将其划分为匹配层和非匹配层,其中划分至匹配层的道路类型数量较少;其次,对匹配层匹配,并只在同种类型道路集中搜索匹配对象,从而避免全局遍历;然后,将非匹配层中剩余未匹配的道路视为新道路网,重新依据拓扑关系分类,划分出新的匹配层和非匹配层,仍按相同方法进行匹配,如此迭代直至匹配结束;最后,对少量无匹配对象道路进行全局遍历检查,作为提高匹配正确率的有效补充。实验结果及对比分析表明,该方法避免了全局遍历匹配,减少了不同层次间道路的干扰,有效提高了匹配的效率和正确率。
In road network matching,matching efficiency is extremely low when using the ergodic matching method that searches matching objects in the entire road network.To improve matching efficiency,a hierarchical and iterative matching method based on road classification is proposed.Firstly,roads are classified based on topology and divided into a matching hierarchy and un-matching hierarchy according to the road type.The number of roads in the matching hierarchy is few;secondly,only roads in the matching hierarchy are matched,avoiding a search for a matching object from the entire road network;thirdly,the remaining unmatched roads are regarded as a new road network so these steps can be repeated.In this way,matching is completed when iteration ends.An ergodic check of the roads that do not have matching objects is done in order to improve matching accuracy.Tests and comparitive analysis indicate that this method can avoid ergodic matching and reduces disturbance caused by roads from the different hierarchies,effictively improving both efficiency and accuracy.