建筑物多边形的多尺度合并化简与平差处理

通过带约束的Delaunay三角形表达建筑物间的空间拓扑关系,提出了取建筑物间Delaunay三角形的最短边同阈值比较,实现了基于SDS（simplicial data structure）模型的建筑物多边形合并的改进算法,及根据最长边最优原则实现合并后对象的直角化处理.探讨了利用直接面状要素综合算法进行建筑物多边形化简,和化简过程中阈值及转折点处理的不确定性带来的综合前后数据质量的变化,由此提出了建筑物多边形合并和化简的整体平差处理模型.算例分析表明,对建筑物合并和化简过程中的平差处理,能提高建筑物在综合过程中的数据质量,并能较好地保持建筑物的形态特征.

Aggregation, Simplification and Adjustment of Building Polygon Objects in Multi-scale Map Generalization

Based on the constrained Delaunay triangulation principles,the neighbor relation of the building polygons is first formulated.An improved algorithm is then presented for the building polygon aggregation on the basis of simplicial data structure(SDS) model,and the principle of rectangular process for the aggregated polygons is proposed.The building polygon simplification is explored by means of the direct area generalization algorithm.Due to the uncertainty in the simplification process with the uncertainty of the threshold value and turn points,a simultaneous adjustment model based on least squares for the building polygon aggregation and simplification is thus established.The case study results illustrate that the adjustment model for the building polygon aggregation and simplification improves the quality of multi-scale cartographic generalization and well maintains building characteristics.