Image watermarking in the Fourier domain using the Mojette transform

The work presented here deals with digital image watermarking and a new Fourier based embedding method is presented. The watermarking location scheme optimization is performed by the use of a psychovisual criteria suitable with the Fourier domain and given from properties of the model of the human visual system (HVS). The watermark design and extraction exploits the Mojette transform properties which are recalled. The embedding as well as the extraction processes are described and exemplified through some classical attacks onto a standard image.