Abstract: The toughest challenge facing the robust watermarking is the geometrical attacks. A slight even intangible geometrical attack can fail the watermark in wavelet domain due to the shift-variance of wavclet. Based on the feature points, this paper proposed an image watermarking scheme against the geometrical attacks. First, according to the tree structure of the wavelet coefficients, the proposed scheme selected the root of the directional subtree with the highest texture as the embedded points from each wavclet tree. Second, this paper proposed an adaptive embedding strategy according to the energy of the low frequency coefficient corresponding to the embedded point and the texture characteristics of the highest frequency coefficients of the subtree. Finally, the proposed scheme used the Harris-Laplace operator to extract the feature points robust to the geometrical attacks,and form a template of feature points. During detection,the proposed scheme restored the attacked image by linear transformation using the feature template, and then verified the watermarking through statistical correlation. The detection process does not need the original host image. Experimental results show that the proposed algorithm has good transparency and is very robust to common image processing and geometric attacks.

Key words: Digital watermarking,Ueometrical attacks,Wavelet tree,Geature points,Template