An Algorithm For Image Digital Watermarking Based On Biorthogonal Wavelet Transform

Nowadays,the rapid development of computer and network have greatly changed people's life. Many inventions and productions are stored or transported by digital ways. And multimedia information exchange based on computer and network also offer the convenience for the usage of digital works. However, digital works prevailing on the net are easily captured,duplicated and juggled , which seriously impinges the copyright of collectivity and personage. In order to solve this problem, as a powerful measure to protect the ownership, digital watermarking is conceived. So-called digital watermarking is a technique of embedding some valuable information about copyright in digital works. Thus, it can protect the ownership. In a total digital watermarking system, the loss of watermarks means the loss of ownerships. Obviously, system like this fails to protect the copyright. So it is necessary that the digital watermarking must possess some characteristics as the followings: 1, Invisibility (or clarity): that is, the original work mustn't be reduced its quality after a series of hiding processings, while the hiding data can't be seen or listened by people. 2, Robustness: It is the most important characteristic of all. It refers to that the whole system can resist some signal processing means and even intended attacks. What's more, the watermark can still be drawn correctly. 3, Certainty: The watermark associating ownership should be exclusively identified. 4, Utility: A good digital watermarking method ought to be suitable for many kinds of files and medias. Consequently, it is practical and can be easily carried out. Wavelet analysis is an active researching field in mathematics and engineering. Comparing to traditional Fourier analysis, wavelet analysis has very good time-frequency character in signal processing. Moreover, it has the fast decomposition and reconstruction algorithm of Mallat. With the theory of wavelet steadily perfected, its applications become widely into many fields such as signal and image processing,pattern recognition,computer vision and graphics.Nevertheless, the symmetry and orthogonality can't be both satisfied at the same time in wavelets. For instance, Daubechies wavelets are not symmetrical. So is to many other orthogonal wavelets. Yet, there are two additional cases: one is the famous Haar wavelet; the other is multiwavelets which are formed by over two scaling functions. Because of the difficulty in constructing orthogonal basis, Albert Cohen, Daubechies etc. put forward the concept of approximate orthogonal wavelet ——biorthogonal wavelet. Biorthogonal wavelets show its excellent quality in image processing due to it considering both the symmetry and orthogonality. The Mallat algorithm for biorthogonal wavelets is: ????? == ∑∑??∈+∈+lZjnjllZjnjldcglncchln(2)(2),1,,1, and c1 ,c,h (n2l)d,g^n2ll ZlZj n= ∑jl?+∑jl?+ ∈∈ In this paper, a method of embedding the image watermarking in a digital picture based on the Daubechies 9/7 biorthogonal wavelettransform is proposed. The followings are just the rough process of it. Step 1: Making use of two-level biorthogonal wavelet transform to host image and selecting the suitable location for embedding the watermark. Step 2: Scrambling and approximating the image watermark to ensure the invisibility of watermarking system. Meanwhile, two independent secret keys produced during the pretreatment can enhance the security of the watermark. Step 3: Using the manner of replacing, we embed the pretreated watermark into the host image which has been transformed by biorthogonal wavelet. Then, reconstructing the host image that has included the watermark, we complete the method. Because of our adopting manner of replacing to embed the hiding information, the host image needn't exist when the watermark is being drawn, which enlarges the extension of application. Anyone who wants to detect the watermark only needs doing these things: Firstly, performing corresponding level wavelet decomposition to the image that is to be detected; drawing the pretreated watermark...