A novel Generalized Arnold Transform-based Zero-Watermarking Scheme

In this paper, based on generalized Arnold Transform (GAT), a novel image zero-watermarking scheme with spread spectrum and despreading (SSD) techniques is presented by using spatial domain of watermark image. GAT is obtained by adding number of variables, and SSD is introduced from communication field to deal with pixels of watermark image. First of all, GAT is used to scramble an original watermark, in which the number of secret keys is increased to improve the security of watermark. Then a binary matrix can be obtained from the original carrier image in quantitative embedding rules. And by using spread spectrum technique, the characteristic matrix can be got from XOR operation between watermark and binary matrix, which is helpful to promote blind extraction of watermark. The embedding process does not change the pixels of carrier, but creates a zero-watermark carrier, which realizes the imperceptibility of watermark and improves the embedding capacity of carrier. In order to achieve the blind extraction of watermark, a binary matrix is calculated from the watermarked carrier by using quantitative rules, and based on despreading technique, a new matrix is obtained from XOR operation between the above binary matrix and the characteristic matrix, which is gained from the embedding watermark. Thus, the original watermark can be extracted by generalized Arnold inverse transformation of the matrix. The experimental results show that the proposed method outperforms many existing algorithms against various attacks, and it can serve as not only robust watermark to protect the digital copyright of works but also fragile watermark to protect the integrity of works.