A feature-based robust digital image watermarking against desynchronization attacks

In this paper, a new content-based image watermarking scheme is proposed. The Harris-Laplace detector is adopted to extract feature points, which can survive a variety of attacks. The local characteristic regions (LCRs) are adaptively constructed based on scale-space theory. Then, the LCRs are mapped to geometrically invariant space by using image normalization technique. Finally, several copies of the digital watermark are embedded into the nonoverlapped LCRs by quantizing the magnitude vectors of discrete Fourier transform (DFT) coefficients. By binding a watermark with LCR, resilience against desynchronization attacks can be readily obtained. Simulation results show that the proposed scheme is invisible and robust against various attacks which includes common signals processing and desynchronization attacks. © 2007 Institute of Automation, Chinese Academy of Sciences.