Statistical analysis of Markov chaotic sequences for watermarking applications
Anastasios Tefas, A. Nikolaidis, Nikos Nikolaidis, V. Solachidis, Sofia Tsekeridou, Ioannis Pitas · 2002
In this paper, statistical analysis of watermarking schemes based on correlation detection is presented. Statistical properties of watermark sequences generated by piecewise-linear Markov maps are exploited, resulting in superior watermark detection reliability. Correlation/spectral properties of such sequences are easily controllable, a fact that reflects on the watermarking system performance.