利用数学形态学提取骨架线的改进算法

Improved algorithm for extracting skeleton line of polygon based on mathematical morphology

地图数字化生产和数字地图编绘中都需要用到提取骨架线算法,如矢量化和等级变换等。提取骨架线的算法较多,基于栅格数据的算法通常采用数学形态学方法,对于矢量数据目前较有代表性的是利用约束Delau-nay三角网提取骨架线。本文给出一种利用数学形态学提取骨架线的改进算法,适用于栅格和矢量数据,并能在矢量化时建立高质量的拓扑关系,还较好地克服了数学形态学方法的缺点——节点畸变和端点内缩。实验表明该算法稳定且行之有效。

Extracting skeleton line of polygon is a frequent operation in map digitization and cartographic compilation, e.g. vectorization and grade transformation. There are many algorithms handling this issue.The rasterbased algorithm usually adopts mathematical morphology; and the typical vectorbased algorithm makes use of constrained delaunay triangulation network recently. This paper proposes an improved algorithm of using mathematical morphology to extract skeleton line of polygon. The algorithm, for raster and vector data, can set up high quality topologic relationship and remove the errors of mathematical morphology. Experiments show that the algorithm is robust and practical.