To achieve the copyright protection of digital image

we present a novel zero-watermarking algorithm based on DT-CWT(dual-tree complex wavelet transform).The algorithm doesn't modify any information of the image

therefore it possesses both the invisibility and robustness

while it can resolve the loss of energy of conventional watermarking algorithms based on redundant transform.Exploiting the idea of real wavelet zerotree structure

we choose the vital-coefficient-trees from the transformed image

then by using PCA(principal component analysis)

extract the first principal components

quantize and encode them to construct the zero-watermark signal

which is registered to the center of authentication and is used to identify user's copyright later.Experiments show

that the robustness of the algorithm is very good

and it can resist various attacks such as filtering

noising

lossy compression

etc..