Enhanced SVD based watermarking with genetic algorithm

Abstract

In this paper, we propose an enhanced robust watermarking algorithm based on singular vector decomposition(SVD). Previous SVD based watermarking algorithms have a problem of false-positive detection. For solving this problem, our algorithm performs an extra verification job for watermark using reference images. To find optimal scaling factors in the algorithm which satisfy both robustness and high quality of the watermarked image, we employed the genetic algorithm(GA). © 2009 Springer-Verlag Berlin Heidelberg.