A Novel Digital Watermarking Based on General Non-Negative Matrix Factorization
Zigang Chen,Lixiang Li,Haipeng Peng,Yuhong Liu,Yixian Yang
DOI: https://doi.org/10.1109/tmm.2018.2794985
IF: 7.3
2018-08-01
IEEE Transactions on Multimedia
Abstract:In this paper, we propose a novel general non-negative matrix factorization (general-NMF)-based digital watermarking scheme for copyright protection and integrity authentication of the image content. Specifically, the proposed general-NMF algorithm is able to factorize a matrix $C\in R_+^{s\times t}$ into a basis matrix $A\in R_+^{m\times n}$ and a coefficient matrix $B\in R_+^{p\times q}$ by removing the dimension-matching constraints required by the conventional NMF, where $s=m$, $n = p$, and $t= q$ . In particular, $s= m\cdot l/n$ , $t=l/p\cdot q$, and the variable $l$ is the least common multiple of $n$ and $p$. Furthermore, the generator factor of the random matrix and $n$ are used as the keys of the proposed digital watermarking scheme. Experimental results show that the proposed digital watermarking scheme can effectively resist various attacks and tampering.
computer science, information systems,telecommunications, software engineering