中文摘要：

矢量线要素数据来源多样,细节层次不一,限制了已有匹配算法正确率的提高,同时也给算法评价带来困难。化简可以减少线要素细节层次,提取其主要形态,据此提出一种基于线要素动态化简的匹配算法评价新方法。对不同匹配算法采用相同数据,在相同化简算法支撑下进行匹配,从而实现对不同匹配算法的评价。首先,阐述动态化简方法提取线要素主要形态的过程;其次,利用动态化简分别辅助4种已有匹配算法,获取每个匹配算法的最优匹配正确率;最后,将4种匹配算法的原始匹配结果与加入动态化简后的匹配结果进行对比,分析化简对匹配结果的影响,并把该影响运用到匹配算法的比较和评价中来。其中,1通过匹配正确率变化、误匹配等分析了匹配算法的数据适用性;2通过化简比例系数K变化时新增匹配数量的统计,评价了匹配算法对线要素局部细节的敏感程度并提出该指标的量化方法;3结合匹配算法采用的匹配相似度指标对其作出评价。

英文摘要：

The vector line feature is characterized by its diverse sources and complicated levels of details, which constrains the improvement of the matching accuracy and the assessment for the existing matching algorithms. Simplification can reduce the levels of details of the line feature, thus to extract its main shape. This paper puts forward a new assessing method for matching algorithm based on dynamic line simplification, in which the assessment of different matching algorithms is realized by respectively combining them with the same simplification algorithm to match the same data. Firstly, describe how to extract the line feature＇s main shape by dynamic simplification; then use simplification to respectively assist four matching algorithms and separately get their optimal matching accuracy. In order to analyze the simplification＇s influence on matching algorithms, the four after-simplify matching results are compared with those corresponding original results, then introduce the above influence into the comparison and assessment. It＇s divided into 3 aspects. Firstly, the adaptability to data is reflected by analyzing the changes of matching accuracy and some wrong matching examples. Then, the statistics for the new-added matching number under different proportional coefficient K in simplification are obtained, through which the sensitiveness to the local details of the four algorithms is assessed, and its quantitative measure index is proposed. Finally, the advantages and disadvantages of the four algorithms are summarized considering the difference of their adopted similarity indexes.