A Novel SVD-Based Zero-Watermarking Scheme for Copyright Protection
Yongchang Chen, Weiyu Yu, Guoming Chen, Qiang Chen, Qian Zhang, Haibo Shen · 2019
In this paper, a novel zero-watermark copyright-protection scheme base on Singular Value Decomposition (SVD) for digital images is proposed, which takes advantage of the essence of SVD, Fuzzy Possibilistic C-Means clustering (FPCM), and image normalization. Image normalization provides robustness against geometric distortions of the host image, and the verification key from the local features are effectively generated by SVD and FPCM. Copyright are then been verified publicly by using digital signature and timestamp. With the verification key, the logo image can be extracted completely without the host image. Compared with relative schemes, the algorithm in the paper is more effective and robust against common image processing and geometric distortions.