Capacity analysis of periodical watermarking

Abstract

Periodical watermark embedding has been especially proposed to cope with geometrical attacks. Using a diversity approach, this method allows to decrease the probability of error in the case of additive attacks. It is usually admitted that the worst additive attack consists in the addition of additive white Gaussian noise (AWGN). However, our theoretical capacity analysis of periodical watermarking demonstrates that periodical AWGN in the optimal attacking strategy leads to more significant drop of the capacity than AWGN test channel from rate distortion.