A Robust Zero-Watermark Algorithm Based on Singular Value Decomposition and Discreet Cosine Transform

A robust zero-watermark algorithm is proposed, which is based on singular value decomposition and discreet cosine transform. The image is firstly spilt into non-overlapping blocks. Afterwards, every block is conducted with singular value decomposition, and its singular value matrix is transformed with discreet cosine transform. The robust zero-watermark sequence is derived from comparing the numerical relationship between two direct coefficients from adjacent blocks. Experimental results of robustness tests show that it has good robustness against various attacks.