Robust watermarking using conjugate symmetric sequency-ordered complex Hadamard transform and normalization

The paper presents a new zero-watermark method which incorporates the normalization process and conjugate symmetric sequency-ordered complex Hadamard transform (CS-SCHT). The normalization process is first performed on the host image to compensate the affine transformation, the watermark is then embedded in the phase component of CS-SCHT domain in a square around an invariant centroid extracted from the normalized version. Experiments show that our scheme is robust to common and affine transformations, and not only provides similar results with DFT, but also the lower computation load than the latter.