An efficient index structure for high dimensional image data

Abstract

The existing multi-dimensional index structures are not adequate for indexing higher-dimensional data sets. Although conceptually they can be extended to higher dimensionalities, they usually require time and space that grow exponentially with the dimensionality. In this paper, we analyze the existing index structures and derive some requirements of an index structure for contentbased image retrieval. We also propose a new structure, called CIR(Contentbased Image Retrieval)-tree, for indexing large amounts of point data in high dimensional space that satisfies the requirements. In order to justify the performance of the proposed structure, we compare the proposed structure with the existing index structures in the various environments. We show through experiments that our proposed structure outperforms the existing structures in terms of retrieval time and storage overhead.