A New Fractal Watermarking Scheme Based on Image Normalization

Geometric attacks are the challenging problems for most image watermarking today. Such attacks are very simple to implement, yet they can defeat most of the existing watermarking algorithms without causing serious perceptual image distortion. In the paper, a new fractal watermarking scheme robust to geometric attacks is proposed based on image normalization and fractal coding. Firstly, the geometrically invariant space is constructed by using image normalization. Then, a significant region is obtained from the normalized image by utilizing the invariant centroid theory. Finally, the watermark is embedded into the self-similar blocks DCT transformed of the significant region by fractal. Experimental results show that the scheme is not only invisible and robust to common signals processing such as median filtering, sharpening, noise adding, JPEG compression, etc, but also robust to the geometric attacks such as affine transform, local geometric distortion, etc.