Font Size: a A A

Study On Digital Image Zero Watermarking Algorithm

Posted on:2013-08-16Degree:MasterType:Thesis
Country:ChinaCandidate:S ChengFull Text:PDF
GTID:2248330371461820Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Digital watermarking is a kind of new technology that can realize information hiding and tracking in an open network environment. The basic idea is to embed copyright information that is on behalf of the owner’s identity into the digital products, in order to hit the target such as copyright protection, content authentication and so on. As digital image information is visual and vivid, it is an important means to express information in our daily life. Therefore, the digital image watermarking technology has become a hot topic.According to different application situation of digital image, this thesis puts forward two types of watermarking algorithms:One is zero watermarking algorithm that will not damage the image quality at all. The other is multi-functional watermarking algorithm that can realize the functions of copyright protection and content authentication at the same time. The main research work and contributions of this thesis are summarized as follows:1. The thesis introduces the basic principle, characteristic and classification of zero watermarking and fragile watermarking technology, makes an analysis and comparison of the difference between the fragile watermarking and the robust watermarking. Introduced separately some typical zero watermarking algorithm, fragile watermarking algorithm and multi-function watermarking algorithm which are based on spatial domain and transformation domain.2. According to the advantages of SUSAN edge detection method such as better detection effect and less computational complexity, the thesis presents a zero watermarking algorithm based on SUSAN edge detection and log polar mapping(LPM).First the proposed algorithm uses the SUSAN edge detector to extract the important edge information of image, then makes use of log polar mapping which can transform rotation, scaling transformation in Cartesian coordinate to cycle translation quantity in LPM coordinate to construct and embed watermarking.Detecting the test image involves watermarking or not by the method of exhaustive search.The experiment results demonstrate the proposed method is not only robust enough to adding noise, filtering and JPEG compression, but also robust enough to various geometric attacks such as rotation, cropping, scaling, and the phenomenon of misjudgment is not easy to happen.3. According to the advantages of curve let transform that have better edge character expression ability, a multi-function image watermarking algorithm based on fast discrete curve let transform is proposed. Taking advantage of the relationship of low frequency coefficients in curve let transform domain to insert the fragile watermarking into the least significant bit in the space domain of the image, on that basis, making use of the image containing fragile watermarking to generate the characteristic matrix of the robust zero watermarking. The experiment results indicate that the fragile watermarking has good positioning function to all kinds of tampering, and the robust watermarking has strong resistance to all kinds of attacks, and realize the functions of copyright protection and content authentication at the same time, overcome the problems such as embedding fragile watermarking and robust watermarking at the same time is easy to damage the quality of image and cause mutual interference between the two kinds of watermarking.
Keywords/Search Tags:digital image watermarking, zero watermarking, multi-function watermarking, edge detection, log polar mapping, fast discrete curve let transform(FDCT)
PDF Full Text Request
Related items