基于特征的等高线数据聚类方法

Data Clustering of Contour Lines Based on Shape Characteristics

针对地理线要素具有的分形特征,将小波分解算法和数值统计算法结合起来,提出并证明等高线小波分解后各层细节分量与复杂度之间具有对数线性的关系,从而计算出等高线的复杂度。再结合等高线自身的位置关系,给出等高线的分类标准,将等高线进行聚类,从而降低图形处理时等高线图形的密度和所需要的等高线的数据量,减少等高线相交的可能性。

Automatic cartography including automatic map generalization and emulational displaying has become an important geographic research area. During this process, it is difficult to solve intersecting of contour lines. For solving that, this paper presents a new clustering method to classify contour lines which is based on geography line factor's fractal character together with its position, combining wavelet analyzing and numerical statistic, proving the logarithm linearity between details of decomposition of contour lines.