Font Size: a A A

Research On Transform Domain Digital Watermarking Algorithm Against Rotation And Scaling Attack

Posted on:2013-03-28Degree:MasterType:Thesis
Country:ChinaCandidate:C J HuangFull Text:PDF
GTID:2248330395968182Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
Robustness against geometric attacks, especially against rotation and scaling, is a difficult problem in transform domain watermarking algorithm, and it is regarded as the bottleneck of the practical application of digital watermarking. The advantages and disadvantages of the existing digital watermarking algorithms have been analyzed. Then two new digital watermarking algorithms have been proposed by combining with the second generation Curvelet transform. The following are the main work has been done.1. The implementation process of fast discrete Curvelet transform and the structure of Curvelet coefficient have been analyzed. The common "scratch" problem in Curvelet transform and its solution have also been discussed.2. A second generation Curvelet transform based robust watermarking algorithm has been proposed. To improve the safety, a random matrix and watermark data XOR were applied. Using multi-layer Curvelet coefficient to embed watermark data, the embedding location can be selected automatically. The numerical experiment results showed that the algorithm has improved the capacity of resisting JPEG compression attack. The algorithm has also met the requirements of watermark invisibility. Its robustness against noise, cropping, filtering and other attacks has also been showed. The algorithm is simple and fast, and its overall performance is better than the wavelet domain algorithm in reference[16].3. Through combination of spatial and frequency domain, a digital watermarking algorithm against geometric attacks has been proposed. It showed the advantages of spatial and frequency domain watermarking. The frequency domain watermark based on the second generation Curvelet transform has been embedded in the green component of color image. The spatial domain watermark has been embedded in the blue component of color image. The frequency domain algorithm uses wrapping method for fast discrete Curvelet transform. The algorithm is simple and fast. The numerical experiment results showed that this algorithm is effective and can better balance the contradiction between the watermark invisibility and robustness. The algorithm can effectively resist geometric attacks (including combination of geometric attacks) and a variety of common signal processing attacks as well.
Keywords/Search Tags:Curvelet, digital watermark, geometric attack, rotation, scaling, robustness
PDF Full Text Request
Related items