Research And Implementation Of Image Watermarking Technology Based On Transformation Domain

With the development of computer and communication technologies, data exchange and transmission has become a relatively simple and quick process. Using electronic installations, people may transmit the digital information rapidly; Carry out academic exchanges and the electronic commerce activities. But the side effect is also obvious, malicious individuals or groups may copy and broadcast the copyrighted content with no owner's permission .So how to solve the problems of copyright protection and the information security becomes an imminent realistic question. Digital watermarking has the dual securities, Embedding watermark is unknown, the key protect the watermark extraction. It provides a new solution idea for the protection of copyright.In this thesis, I study digital watermarking of images based on transformation domain. I have read a large amount of documents and materials about digital watermarking technology, and then I sum up the developing history of multimedia digital watermarking, frame models etc. After the conception, specialties and application domains of digital watermarking are introduced, many kinds of embedding technology of digital watermarking is analyzed carefully, and point out the advantages of algorithm based on the transformation domain. Image watermarking algorithm can be roughly divided into two categories: the space watermarking and transformation domain watermarking. Most of the space domain watermarking algorithm lack of stability of the image processing. But transformation domain watermarking algorithm has many advantages: watermark energy can be spread to all the pixel space domain, this will help to ensure that the invisible watermark in the transform domain; the human visual system can more easily integrated into watermark encoding process ; it compatible with the international data compression standard. Transformation domain algorithms are highly robust.This thesis proposes two watermarking algorithm .One is a DCT domain image digital watermarking algorithm based on Arnold transformation. The algorithm embeds watermark in the low-frequency coefficients use multiplication rules. Before embedding, we use Arnold transform, even watermark was extracted, and it is difficult to restore the original watermark image. After scrambling, the watermark image pixel distribution to the entire space, so even if the watermark image suffered malicious attacks, we can extract good watermark. Another algorithm is based on the mixed transform domain of DCT and DWT. First, we make wavelet decomposition for the primitive image; wavelet approximation graph is the ideal regional of watermark embedding. But it is the human visual sensitivity, so we make DCT transformation, and watermark is embedded in the DCT-IF coefficient. And carried on a lot of robust testing laboratories for the two algorithms, through these experiments analyzed the anti-attack capability of each algorithm regarding the JPEG compression, the value filter, the noise, cutting and revolving. This article research content is the international related domain the front topic, has the important theory significance and the application prospect.