Bulk Construction Algorithm of Clustered Sorting Generics 3DR-Tree
Zhenwen He · Geography and Geo-Information Science · 2009
A new 3DR-Tree construction techniques for 3D spatial index is proposed.The main ideal of this algorithm is to make the spatial objects that near to each other in spatial space in nearest leaf nodes,and to reduce the overlap among the spatial objects' rectangles.Given a collection of multi-dimensional spatial objects with rectangles,all the spatial objects will be clustered them to K groups by distance relativity,and be sorted in the i-th(i∈) group,and then all the groups will be sorted by the group center points,and the R-Tree is built bottom-up at last.The experimental results show that the 3DR-Tree constructed by this method outperforms the previously R-Tree methods in query efficiency and space utilization.