The Research Of Adaptive Watermarking Hiding Algorithm Based On Human Visual System

The rapid development of the computer and the Internet technology allow people to storage,copy and spread the digital products on the Internet even though they have no copyright.Therefore,it is a serious threat to digital product security.Digital watermarking technology can protect the rights and interests of the copyright owner,so its development has attracted wide attention.Among variety of watermarking technology,the watermarking technology of digital color image gains focus in this research field.By analyzing the existing algorithms,this thesis proposes two new algorithms on double color image watermarking,whose main idea is the adaptive embedding.At the same time,the robustness,invisibility and security of the algorithm are verified.The main work and novelty are summarized as follows.Firstly,upon summarizing the development of digital watermarking technology and analyzing the research status,this thesis focas its research topic on double color image watermarking algorithm.In order to reduce the impact on the host image for embedding watermark image,the adaptive masking value of the color host image is constructed based on the masking properties of the human visual system about the texture,luminance and edge of the color host image.Thus the digital watermarking image is adaptively embedded into the host image.For texture masking value of image,this thesis utilizes a new calculation method based on spectral measurement.The method is proved to be efficent on identifying the global texture both in the frequency-domain and spatial-domain.Secondly,combining Nonsample Contourlet Transform(NSCT)and Human Visual System(HVS),this thesis proposes an adaptive color image watermarking algorithm by exploiting dual characteristic of time-frequency of Fractional Fourier Transform(FRFT).At first,the maximum embedding strength that the color host image parts can tolerate is determined by calculating the adaptive masking value of the color host image.Then by using the additive principle of embedding,the color watermark is embedded by modifying the FRFT intermediate frequency coefficients of the color host image.Numerical results show that the algorithm has three advantages: better invisibility,higher robustness and higher security.Thirdly,in order to improve the invisibility and robustness of the color watermark image,this thesis proposes a new blind double color image watermarking algorithm based on HVS and Schur decomposition.At first,the watermark embedding position is determined by using the improved method,which is used to calculate the adaptive masking value of the color host image.Then the unitary matrix is obtained by Schur decomposition.By utilizing the stabilizing analogy of the elements in decomposed matrix,the elements of the matrix can be changed accordingly.Thus,not only the embed watermark is fulfilled,but also the watermark is used to blind extract.Experimental results show that the proposed algorithm has better invisibility,and it is robust to image attacks,such as salt and peppers noise,low-pass filtering,sharpening,and blurring.