Abstract: Clustering of polygons is an important means of mining the intrinsic spatial knowledge of polygons. The current problems of varying size, morphology and distribution of polygons lead to less accurate clustering results. At the same time, in order to meet the needs of analysis of large batch of polygons data,this paper proposes a clustering method for polygons with multi-density attribute calculation index. Firstly, according to the different locations (including boundaries) of the internal points of a single polygon, multiple density attributes are assigned to a single polygon. Secondly, based on the tendency of the low density values among the polygons for converge to the high density values, a polygon aim vector is generated, and the tree structure connections of the elements are sequentially generated. Finally, the clustering of polygons is completed by the strategy of connection pruning and merging. It is proved that the method can effectively identify the aggregation clusters of various irregular polygons, and has good accuracy performance in the aggregation of massive polygons data, realizing the clustering needs of recognition of high density polygonal regions.

Key words: big data, polygons data, cluster analysis, polygons clustering, multi-density attribute