Rotation, scale and translation invariant spread spectrum digital image watermarking

Abstract

A digital watermark is an invisible mark embedded in a digital image which may be used for Copyright Protection. This paper describes how Fourier-Mellin transform-based invariants can be used for digital image watermarking. The embedded marks are designed to be una ected by any combination of rotation, scale and translation transformations. The original image is not required for extracting the embedded mark.

DHT domain digital watermarking with low loss in image informations

“…The change in the pixel values, due to watermark embedding, can be obtained by subtracting Eq. (3) from (6) and is expressed as follows:…”

“…Several digital image watermarking schemes using discrete Fourier transform (DFT) [6], Fourier-Mellin [7], discrete cosine transform (DCT) [8,9], and wavelet transformations [9,10] have already been reported in the literature. Among them, DCT and wavelet transforms become popular as the most common image compression techniques e.g.…”

Robustness and security of digital watermarks

“…If a transformed representation is used, it may be applied to the entire data (the whole image), or on a block-by-block basis. For images, watermarks have been applied to pixels [2], Fourier components [24], whole-image and block-based discrete cosine components [8,23,39], wavelet components [23], and Fourier-Mellin components [25,13]. One use of a transformed representation is to make the components of the data more independent; pixels in an image, for example, are highly correlated locally, which is not true of discrete cosine components.…”

“…For example, one may choose to represent a color image using a luminosity-chrominance basis, and mark only the luminosity components, thereby rendering the watermark robust to a color-to-grayscale transformation. A similar idea discussed more fully in the next section is to mark certain Fourier-Mellin components of an image [25,13], thereby obtaining some robustness to the geometric transformations of scaling, cropping and rotation. To enhance security, one may choose to mark only a random subset of the set of components selected for robustness.…”

“…6 Ruanaidh and Pun [25] have explored Method 4, that of marking the magnitudes of the Fourier-Mellin components, as a way to make watermarks robust against translation, rotation and scaling. This idea has been refined by Herigal, et al [13].…”