建筑物白模多边形的自动合并

Automatic Aggregation of Building Footprint Polygons

  • 摘要: 建筑物白模多边形数据可广泛应用于许多领域,但在实际应用中,由于数据太过细致,且目前使用的建筑物白模多边形数据存在拓扑关系错误,不满足生产要求,这给地图综合中建筑物群的自动合并提出了新的要求。因此提出了一种基于约束性Delaunay三角网的建筑物白模多边形自动合并方法,在保持建筑物整体结构和视觉效果的前提下减少不必要的细节。该方法能够消除原始建筑物图形数据中的空间拓扑关系错误,并对建筑物白模多边形进行一致性处理,有利于独栋整体建筑物的三维可视化。并用C#编程语言开发桌面应用程序对该方法进行了实现,通过和ArcGIS建筑物综合工具的建筑物群合并结果进行对比分析,验证了该方法的有效性和实用性。

     

    Abstract: Building footprint polygons have been used in a wide range of aspects in many fields. However, they are too detailed in many applications to meet the production requirements. Aggregating building footprint polygons can remove unnecessary details while preserving the overall structure and visual impression of them. Nowadays they are usually generated from high resolution cadastral and remote sensing data, having many topological problems. These topological problems will influence application of the data, then bring new requirements to buildings aggregation in automated cartographic generalization. So we should solve these new problems. This paper proposes a method to aggregate building footprint polygons automatically, which can remove these topological problems. This method can effectively maintain the consistency of spatial topological relationship among building footprint polygons, and make three dimension visualization of independent buildings more convenient.And a desktop application to implement our proposed method in this paper. Comparison and analysis of experimental results with results from polygon generalization tool of ArcGIS verified that the proposed method is effective and practicable. The results show that our method has following advantages: (1) This method can remove topological problems while aggregating building footprint polygons. (2) Our classification of triangles is beneficial for aggregation of building footprint polygons. (3) This method can acquire overall building contour conforming cognitive knowledges and independent building footprint polygons with topological consistency at the same time.

     

/

返回文章
返回