扩展分维在地图信息综合中的应用

Application of extended fractal dimension in map generalization

文中强调通常的分维与扩展分维的本质差异。作者主张综合应遵循非自相似准则,即非线性原理。文中用D-P方法建模,因为后者所使用的限差ε直接和地图曲线信息综合的D-P方法相联系。关键问题是要将多分辨率序列与多比例尺序列实现某种对应。本文用基于河流自身的自动预综合机理分析方法建立分维模型参量(L)序列与地图信息综合主要参量(M)序列之间的映射关系,最终把地图比例尺序列Mi与粗视化限差序列iε联系起来,借此实现基于扩展分维的河流自动综合。

In this paper the essential differences between the ordinary fractal dimension and the extended one were emphasized. Author had the point of view that the generalization should follow the non-similarity principle, namely the nonlinear principle. Extend- ed fractal modeling was implemented by D-P method, because it uses the threshold 8 which has a direct relationship with the D-P method of curve generalization. The key problem is to establish the rational correspondence between the multi-resolution and multi-scale series. Author established the mapping relation between fractal dimension parameter L and map generalization parameter M through the automatic pre-generalization of river entity. Thereby the extended fractal based river feature generalization was realized in principle.