Abstract:

In order to solve the problem that close military symbols may shelter each other while reducing scale of map in computer plotting, by analyzing the parameters of neighborhood, this paper uses Density-Based Spatial Clustering of Applications with Noise(DBSCAN) algorithm to seek the symbols sheltering each other, and uses plot instead of symbols in their centroid. Aiming at the shortcomings of DBSCAN algorithm, it proposes an improved algorithm named Based on Density and Irregular Region Clustering of Applications with Noise(BDIRCAN), which considers application conditions and changes the traditional circular neighborhood to the applied irregular polygonal neighborhood. Experimental result shows that BDIRCAN can solve the problem well and avoid clustering the symbols which stay near but do not shelter each other.

Key words: Density-Based Spatial Clustering of Applications with Noise(DBSCAN) algorithm, radial algorithm, clustering, plot, symbol

摘要：

为解决计算机标图过程中因缩小地图比例尺而导致的标号扎堆问题，通过分析邻域参数，利用DBSCAN算法寻找相互遮挡的标号，在其质心处用标图代替扎堆标号。针对DBSCAN算法的不足，结合实际应用情况，将传统基于密度方法的圆形邻域改为针对应用的多边形邻域，提出改进的算法BDIRCAN。实验结果表明，BDIRCAN算法能较好地解决标号扎堆问题，避免对临近但不相互遮挡的标号进行错误的聚类。

关键词: DBSCAN算法, 引射线法, 聚类, 标图, 标号

CLC Number: 

• TP301.6