城市地图数字合并变换算法的比较研究

Comparison study of merging algorithms in digital map conflation

地图数字合并是实现不同来源数据融合和集成的方法和技术。匹配与调整变换是数字合并的两个主要过程,匹配过程存在大量的不确定性,使得一些实体无法找到匹配实体。因而在调整变换过程中,选用合适的方法以达到最优效果尤其重要。从定性与定量分析的角度,确定了定性与定量的检验标准,并采取一定实验手段,对城市地图数字合并采用的调整变换方法进行比较分析,依据分析结果,得出适用于城市地图数字合并的调整变换方法。

Map conflation is a methodology to remove discrepancies generating from overlaying two datasets to get a new and consistent dataset. The two phases of map conflation, feature matching and merging, use their respective algorithms. Due to the uncertainty during the matching procedure, some features cannot be correctly matched. Therefore feature merging is needed, which can eliminate inconsistency between two source maps. Several algorithms have been applied in map merging phase, but they have not been systematically compared. Therefore, it is vital to determine the most suitable merging method to attain the best results. It presents some comparing measures, including quantitative and qualitative ones. Through test it analyzes three main algorithms to determine which is the best to be used in map conflation in city area.