摘要
高维空间中点数据的索引及检索是基于内容图像检索领域的关键问题,文中将LSH(localitysensitivehashing)索引算法应用于基于内容图像检索系统中,与传统的索引方法相比,该算法具有复杂度比较低、支持非常高的维数、极低的I/O代价等特点。实验结果证明,将该索引算法应用于基于内容图像检索系统中,其性能优于传统的索引方法.
It is a critical issue for indexing and retrieval of high dimensional point data in content-based image re-trieval field.In this paper,a new kind of indexing structure is adopted in the content-based image retrieval system,in comparison with traditional indexing methods,the LSH can build with low complexity,support very high dimensionality,and even very low I /O cost,etc.
出处
《计算机工程与应用》
CSCD
北大核心
2002年第24期20-21,63,共3页
Computer Engineering and Applications
基金
国家863高科技发展计划资助项目(编号:863-306-ZD11-03-3)
