Robust watermarking and its applications to communication problems

Abstract

Digital watermarking has recently gained an intense interest in research and applications. An invisible and secret signal, called watermark, is added to the host data. With the help of this watermark issuer of the data can be unveiled, unauthorised users can be identified, illicit copying can be avoided, any attempt to temper with the data can be detected and many other security services can be provided. In this thesis, the relations and differences between watermarking and communication systems are elaborated. Based on these results new methods for both watermarking and communication are derived. A new blind, robust and reversible watermarking scheme based on Code Division Multiple Access (CDMA) is presented in this thesis. Using this scheme watermark is arithmetically added to spatial domain or frequency domain. Watermark is extracted by using spreading codes only. Proposed watermarking scheme is simple, computationally efficient and can be applied to any image format. A novel idea that watermark can be part of the image is presented. By using watermark, which is a part of an image, digital watermarking can be used beyond simple security tasks. A part of an image is selected and embedded in the whole image as watermark. This watermarked image is attacked (transmitted or compressed). By using the extracted watermark and attacked selected part image quality can be assessed or jpeg quantization ratio can be estimated or even image can be equalized blindly. Furthermore, CDMA based watermarking is used to authenticate radio frequency signal. Spreaded watermark is added in the form of noise to the modulated radio frequency signal. If this noise is increased, watermarked signal automatically becomes a scrambled signal. Later watermark is extracted and by using reversibility of proposed scheme watermark is removed. Once the watermarked is removed original signal is restored, hence descrambled.