A simple method for updating the eigenvectors and eigenvalues of a covariance matrix when a new input sample is added is presented. This proposed method will be a solution for both rank-one modification problems of a symmetric matrix and adaptive principal component analysis.
© 2006 Optical Society of America
Original Manuscript: March 13, 2006
Revised Manuscript: September 10, 2006
Manuscript Accepted: October 12, 2006
Published: December 13, 2006
Yongkyu Kim, "Incremental principal component analysis for image processing," Opt. Lett. 32, 32-34 (2007)