基于点密度分析与自适应差异检测的点群要素制图综合算法

An Algorithm of Cartographical Generalization for Point Cluster Features Based on Point Density Analysis and Adaptive Difference Detection

针对点群要素制图综合的自动选取问题,提出一种基于点密度分析与自适应差异检测的点群要素制图综合算法。该算法首先利用核密度估计法对原始点群进行点密度分析,选取凸包最外侧点和重要性等级较高的点分析其点密度,进而对比两次点密度分析结果的相对密度差异,并找出相对密度差异最大的点加入选中点群中;然后重新对选中点群进行点密度分析,直到选中点数满足制图综合要求。将该算法用于实例点群数据中进行了验证,结果表明,该算法能最大限度地保持点群的分布特征和密度对比关系,并且较为简单、灵活。

Aiming at the automatic selection of points in cartographic generalization for point cluster features,an algorithm of cartographic generalization for point cluster features based on point density analysis and adaptive difference detection is proposed in this paper.In the algorithm,the densities of all the points in the original point cluster are firstly analyzed by the method of kernel density estimation.Then,the outermost points of the convex hull and the points with the higher importance level are selected,and the densities of these selected points are analyzed.After comparing the relative density difference of the two point density analysis results,the point with the biggest relative density difference is found out and added to the selected point cluster,and then the densities of all the points in the selected point cluster are reanalyzed until the number of the selected points meets the requirement of the cartographic generalization.The algorithm is also validated in the instance point cluster data.The result shows that the algorithm can keep the distribution characteristics and density contrast relationship of the point cluster to the maximum extent with simplicity and flexibility.