This paper proposes a robust and blind watermarking algorithm based on matrix Singular Value Decomposition (SVD) perturbation theory for JPEG image copyright protection. First, the quantized DCT coefficient blocks are extracted from the JPEG image, then SVD is performed on each block, and finally the watermark is embedded into a set of selected maximum singular values of the quantized Discrete Cosine Transform (DCT) coefficients using dither modulation. From the embedding principle of the watermarking algorithm, we can see that the watermark information can be completely extracted correctly when the watermarked image under certain attack. A large number of experiments show that the algorithm has a higher degree of robustness against attacks such as Gaussian noise and JPEG compression, etc.

/pdf/a-robust-watermarking-algorithm-for-jpeg-images-2kspzilo7z.pdf

Daofu Gong

Papers

A Robust Watermarking Algorithm For JPEG Images