Updating algorithm for extracting the basis of karhunen-loeve transform in non-zero mean data

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. So the efficient updating algorithms were proposed in signal processing and numerical linear algebra. The updating algorithms make the active learning and recognition possible. But they mainly deal with zero mean data. In this paper, we propose an updating algorithm for KLT for non-zero mean data. And we also show its application in face analysis. The experimental results demonstrate the efficiency of our algorithm. Karhunen-Loeve transform, non-zero mean, rank-k updating, singular value decomposition.