Radon transform based robust non-blind watermarking

The purpose of digital watermarking is to attempt to solve the copyright ownership problem for all multimedia data which can be image, audio or video. The main objective of this thesis is to design a robust method of image watermarking for copy protection and copyright protection.The thesis consists principally from an overview of watermarking algorithm, types of watermarking, watermarking requirements. Further the watermarking techniques that are used in embedding. There are many watermarking techniques discussed in spatial domain and transform domains. The spatial domain watermarking algorithms still not resistant to lossy compression image, shed noise in the image and most operations of image processing. On the other hand, frequency domain watermarking techniques can embed more bits for watermark and the watermarked images are more robust to attack. Otherwise, in frequency domain we can embed more bits and more resistance to the attacks. However, comparing watermarking in DWT with watermarking in DCT, DWT has a number of advantages that led us to focus in discrete wavelet transform from other types of same domain.In this thesis, there are three different watermarking methods proposed in different levels DWT based on radon transform. All proposed methods are requires the presence of original image for watermark extraction. Each method is various from the other from the way of embedding the watermark, the number of watermarks that embedded and in which level the watermark embedded. Moreover, the watermarked image and attacked images measured by the PSNR that measuring the noise ratio in each image. Finally, we attempt to extract the watermark and attacked image, to determine the proportion of the deterioration that occurred after the attack or the embedding process.Keywords: Watermarking, Discrete Wavelet Transform, Radon Transform.