土地利用数据综合中的聚合与融合

针对土地利用图中多边形地块具有全覆盖、无重叠、语义上多层次的特点,对土地利用数据综合中的地块合并给出了多边形聚合与多边形融合两种操作;前者针对具有同属性的语义邻近地块的合并,后者针对不同属性的拓扑邻近地块的合并;两种操作的区分是基于同时顾及空间、语义特征的邻近分析,算法的实现均建立在Delaunay三角网模型上由骨架线支持;详细讨论了两种操作的算法过程并给出了实际数据的实验结果。

Aggregation and Amalgamation in Land-use Data Generalization

The land_use map is a typical kind of categorical theme and has the properties covering the whole area with neither gap nor overlapping area.Its generalization has to consider geometric simplification,semantic hierarchy abstraction and statistical properties maintenance.This paper focuses on the combination of neighbor parcels giving two operations:aggregation and amalgamation.The previous aims at the polygon objects with homogeneous semantics,and the latter aims at those with heterogeneous semantics.The decision of neighborhood relationship is the basis of two operations.This study considers both spatial and semantic facts in the judgment of neighborhood cognition based on Voronoi diagram model.The work flow of two operations is presented in land_use data generatization with the experiment result of real data.