Watermark Singular-Values Encryption and Embedding in the Frequency Domain

Abstract

Protecting digital assets has become not only a taken-for-granted practice in most electronic applications for content and ownership authentication, but robustness and efficiency are becoming decisive attributes when exercising watermarking. A highly efficient approach is suggested and which integrates encryption to a hybrid watermarking algorithm based on the mathematical tools of the well-known singular vector decomposition (SVD) and the discrete wavelet transform (DWT). The innovative attribute of this approach stems from the fact that the suggested method achieves the encryption of the data prior to watermarking in the frequency domain. The encryption algorithm is based on a random key source which entitles the user to have a selected key-length and guarantee a significantly advanced level of intruders’ unawareness. The encryption program is applied to the singular values of the watermark. The watermarking algorithm is based on the combination of the SVD and DWT. This watermarking method has high robustness against geometric transformations but not efficient for malign attacks. The combination of encryption and SVD-DWT algorithms overcomes this last weakness while keeping all advantages. As shown by the simulation results, a high level of robustness and security is attained in the sense that the concealed watermark is indestructible and almost undetectable in many cases. Moreover, the capacity of insertion is sustained at a satisfactory level while securing a high signal-to-noise ratio.