为进一步提高R—tree的空间索引数据检索性能,提出一种基于R—tree的MRD-tree索引,它的结构中包含空间对象的外部近似图形和内部近似图形,同时允许数据入口存放在中间结点。在检索时既能减少I/O次数,也能相对减少一定的CPU时间。通过实验分析,该索引能提高空间数据的检索效率。
To further enhance the R-tree spatial index data retrieval performance, a MRD-tree index is presented based on R-tree, which contains both external approximate shape and internal approximate shape of spatial object, and it is allowed that the data entry stores in middle nodes. The number of I/O and the cost time of CPU are reduced during the search by the index, and a series of tests indicate that the MRD-tree index improves the query efficiency of spatial data.