Digital Watermarking based on Fractal Coding in Wavelet Domain

Yang Li,Si-Dan Du

DOI: https://doi.org/10.3321/j.issn:0469-5097.2006.04.007

2006-01-01

Abstract:Digital watermarking technology has attracted a great deal of attention in recent years,because watermarking is good way to protect the intellectual property rights of digital media. The information carried by the watermark can be accessed by using a detection algorithm with the secret key provided. An important property of a watermarking is its robustness with respect to image distortions.This means that the watermark should be readable from images that undergo common image processing operations, such as filtering, lossy compression, noise adding, histogram manipulation, and various geometrical transformations. The fractal theory has proved to be suitable in the field of image coding and watermarking. In general terms, a fractal coder exploits the spatial redundancy within the images by establishing a relationship among its similar parts. It is assumed that many objects can be closely approximated by self-similarity objects that might be generated by use of Iterated Function System (IFS) transformations. According this assumption, the IFS can be seen as a relationship between the whole image and its parts. Thus the main problem is how to find a set of transformations to describe the relationship, in other words, how to define the IFS. Employing the invariant parameters of the fractal transformation, the watermark embedded into the original image can be extracted form the retrieved image. But mosaic effect and bad image quality can be brought by fractal transformation in spatial domain.A novel digital image watermarking method based on fractal coding-decoding and wavelet transformation is presented in this paper. A fractal coding-decoding algorithm in the coefficients of detail images in wavelet domain is presented and studied. On the basis of it, a watermarking scheme is proposed. The watermarking method uses the fractal code which is generated by calculating an Iterated Function System from the detail image. Employing the invariance of the Iterated Function System code,the mark can be hidden and retrieved. In order to greatly improve the robustness and image quality,certain resolution level coefficients in wavelet domain is coded and watermark is embedded into it. The algorithm adds artificial and visually invisible local similarity into the detail image in wavelet domain.This is done by substitution of the range block with a new Matching Block in stead of the original best one. In other words, the watermark is embedded altering the original collage map with a new one which is controlled by the watermark. The proposed watermark embedding consists of a coding-decoding process and the mark retrieving will be performed as a fractal coding process too. Experiments have been performed in order to measure the robustness of the scheme against image processing attack. The watermark is readable from images that undergo common image processing operations, such as filtering,lossy compression, noise adding, histogram manipulation, and various geometrical transformations. In each case, very promising results have been obtained.