All Published Papers Search Service

Title

Sequential Updating Algorithms For Extracting The Basis Of Karhunen-Loeve Transform

Author

Yanyun Qu, Nanning Zheng, Zejian Yuan, Cuihua Li

Vol. 6  No. 1  pp. 24~29

Abstract

Karhunen-Loeve transform (KLT) is a popular method for dimensional reduction and feature extraction in image analysis, signal processing and automatic control systems, and so on. The drawback of the KLT is expensive computation. In this paper, firstly we propose a novel updating algorithm for KLT processing k data at a time, which has advantages especially for image sequences: it is faster than batch algorithm, and can handle the dynamic database, and doesn't save the entire database. Secondly we offer two approaches to solve the ''non-zero-mean"" case, which few works discuss. And finally we analyze the error of the algorithms. We show applications in face analysis. The experimental results demonstrate the efficiency of our algorithms.

Keywords

Karhunen-Loeve transform, rank-one updating algorithm, singular value decomposition, sequential updating

URL