基于非均匀划分的分布式影像数据索引方法

本文介绍了现有的对影像数据管理和索引的方法及其存在的不足，提出了一种基于非均匀划分思想对分布式影像数据进行管理和索引的方法，避免了由于边缘补零带来的时间和空间开销。实验结果表明，本文方法能够支持对分布式多图幅的联合查询以及结果的无缝拼接，而且其时间和空间代价远小于现有方法。

A Method of Distributed Image Indexing Based on Non-Uniform Partitioning

In this paper, the classical managing and indexing methods of image data are presented, the deficiencies of the method are also analyzed. To solve the problems in the classical method, a new managing and indexing method based on non-uniform partitioning is proposed in this paper. It avoids the massive time and space cost caused by zero-padding on the edges of images. The result of experiments shows that the proposed method can support region query and seamless mosaic among distributed images efficiently. The time and space cost is much less than that of the classical ones.