关键词: 虚拟旅游, 3D点云数据, 差值压缩, 动态索引, R树

Abstract: 3D point cloud data in virtual tourism models is very large, the effective compression and spatial index can reduce the transmission bandwidth and query time. Based on the local similarity principle, progressive neighbor data difference compression method is presented. That is, point cloud data spatial subdivision, morton code sorting, data blocking, calculating difference, compressing by a block,the method not only improved the compression rate of point cloud data, but also transmited it in a streaming manner,it was solved that the matter of R-Tree’s index performance degraded owing to multiple queries caused by Overlap of it’s sibling nodes in three-dimensional space. Redundant processing technology of cuting overlap area is puted forward according to the fact that crossover of two adjacent roaming query windows is of high probability. Spatial arithmetic calculating valid range of this query, the method can effectively reduce i/o expenses , redundancy queries, improve query efficiency.Experiments show that the compression and spatial indexing methods have obvious advantages.

Key words: Virtual tourism, 3D point cloud data, difference compression, dynamic indexing, R-tree.