Abstract:

Traditional transform domain watermarking algorithms always embed watermarking signals by modifying transform domain coefficients, thus affecting the imperceptibility of the image. Using the stability of the numerical relationship between the entire image mean and block mean, we propose a robust zerowatermarking algorithm for spatial domain color images. A feature matrix is produced by judging the numerical relationship between the mean of the entire image and each block. The zerowatermarking information is constructed by performing XOR operation on both the feature matrix and the preprocessed watermarking information. After preprocessing, it is registered in the intellectual property rights database. When extracting watermarking, we obtain the final extracted watermarking information by the vote strategy. A large number of experimental results show that the proposed algorithm has strong robustness against conventional signal processing attacks and geometric attacks, such as row and column shifting, scaling and rotation. Compared with similar robust watermarking algorithms, the proposed algorithm has a better performance against most attacks.

Key words: color image, zerowatermarking, spatial domain, robustness