中文摘要：

研究多尺度地理空间点群目标相似关系的计算问题。基于4类点群信息,将点群目标包含的信息重新分成统计信息、专题信息、拓扑信息、几何信息和度量信息,这5类信息是影响点群相似性程度的主要因子。根据这5类信息描述参数的特点,给出了各个因子的相似度计算公式,建立了多尺度地理空间点群目标相似关系的计算模型,并通过算例验证了该模型的可行性与有效性,为地图综合质量的评价提供了一种新方法。

英文摘要：

In this paper,the calculation problem of similarity relationships for point objects in multi-scale geographical space was studied.Based on the four types of information,point cluster′s information was divided into five categories which are statistical,thematic,topological,geometrical and metric information respectively.And these five types of information are the main factors effecting similarity degree between point clusters.According to features of describing parameters of these five types of information,computation formulas of similarity degree for various factors were got and a calculation model of similarity degree for multi-scale point objects was constructed. The calculation results of an example validated that this calculation model is feasible and effective. And a new method was provided for the calculation of the result in map generalization.