The Research And Implementation Of Digital Watermarking Algorithm Based On Wavelet Transform

In recent years,with the development of the digital media industry,the way of information transfer has changed.Previously,people used paper to pass information,and now use digital media to convey information.Digital media makes people's life more convenient,while at the same time,Information security issues become increasingly serious.Originators' rights and publishers' reputation are affected by the condition that infringement behaviors.So there has been many technology for copyright protection,digital watermark technology has drawn the attention of researchers.Digital watermarking technology is an interdisciplinary field involving information security,cryptography,which is an effective means to protect copyright.After analyzed large papers on digital watermarking technology,the following work is completed in dissertation.(1)This thesis researched background and research status at home and abroad on digital watermark technology.The basic frame,characteristic and classification of digital watermark are described in detail.And focuses on the pretreatment technology and typical attack techniques of watermark technology.Analyzed the shortcomings of the traditional algorithms.(2)This thesis not only studied the basic principles and properties of wavelet analysis theory,but also studied the wavelet decomposition and reconstruction of two-dimensional images and its application in digital watermarking.In addition,studied the basic principles,properties and algorithm application direction of Chaos algorithm and SVD.Based on these theories,a new chaotic scrambling formula is provided,which has a specific form and does not contain any random factors,but can produce seemingly completely random sequences.As the pretreatment technology for the improved watermark algorithm.(3)Introduced a digital watermarking algorithm based on chaos scrambling and DWT,which combines new preprocessing techniques with wavelet analysis theory.In this method,the watermark image is decomposed by DWT,the carrier image is decomposed by MRA.And then,in the process of embedding the watermark,different frequencies use different embedding intensities.(4)An improved watermarking algorithm based on singular value decomposition is proposed,because the original algorithm is not robust to the rotation attack.The algorithm is based on the new preprocessing technique and DWT,combined with rotation invariant properties of SVD.It's to do the same wavelet decomposition of the watermark image and the carrier image,and take the low frequency component to do SVD decomposition.Then embeds the S matrix of the watermark image into the S matrix of the carrier image.(5)This thesis simulates the original algorithm and the improved algorithm,and gives the experimental results and analysis.This paper used several common image attack processing methods(such as noise,rotation,filtering,JPEG compression,cropping,etc.)to test the robustness of the algorithm,besides,used PSNR and NC to evaluate the performance of the algorithm.The experimental results show that the improved algorithm has successfully solved the problem,and has good watermark invisibility,robustness and security.