<title>Geometric distortion correction in image watermarking</title>

Abstract

In this paper we report a novel method to estimate the scaling factor of a previously scaled watermarked image and the angle by which the image has been rotated. Scaling and rotation performed on a watermarked image, as part of the attacks the image may undergo, can very easily confuse the decoder unless it rescales and/or rotates the image back to its original size/orientation, i.e., recover the lost synchronism. To be able do so, the decoder needs to know by how much the image has been scaled and rotated, i.e., needs to know both the scaling factor and the rotation angle. In our approach, we compute the Edges Standard Deviation Ratio (ESDR) which gives us an accurate estimate for the scaling factor. The rotation angle is approximated by the Average Edges Angles Difference (AEAD). Both ESDR an AEAD are computed from wavelet maxima locations which have been estimated form the non orthogonal dyadic wavelet transform. The proposed scheme does not require the original image provided that a proper normalization has been attained. Our method has proved its robustness to wide rotation and scale ranges.