A Study On The Theory And Several Issues Of Digital Image Watermarking Technology

With the rapid development of Internet and multimedia technology,people can easily access the multimedia resource what is needed.However,such easy access makes the multimedia resource easier to be illegally copied,tampered and transmitted,which brings serious threats on protection of the lawful right and interests of the copyright owner.As an active authentication method,the digital image watermarking technology has become an important measure to protect the copyright of digital images.However,the existing algorithms are difficult to resist geometric attacks and achieve good balance between the robustness and imperceptibility.In this paper,we propose three effective digital image watermarking algorithms to deal with these problems,which can be summarized as follows:1?A fast computation method of Radial Harmonic Fourier Moments(RHFMs)is proposed,and an image watermarking approach is presented using the fast RHFMs based on geometric correction.Firstly,the Nonsubsampled Shearlet Transform(NSST)is performance on the original host image,and the watermark is embedded into the host image by adaptively modulating lowpass NSST coefficients.Then,the excellent RHFMs and Fuzzy Support Vector Machine(FSVM)are exploited to train the geometric correction model.Finally,the geometric correction is performed on the watermarked image with the well trained model,and the watermark is extracted from the NSST low pass sub-band of corrected watermarked image.In this algorithm,the fast computation of RHFMs is presented,which can improve the ability to describe image features.The geometric correction model is exploited based on the fast RHFMs and FSVM,which can make the watermarking algorithm has a strong ability to resist geometric attacks.Experimental results show that the proposed approach provides good imperceptibility and robustness against common image processing operations and geometric distortions.2?By combining the color invariant model and the correlation between color channels,we propose a local Quaternion Polar Harmonic Transform(QPHT)based robust color image watermarking algorithm.Firstly,the stable color image feature points are extracted by using new point detector,in which the Scale Invariant Feature with Error Resilience(SIFER)detector and color invariant model are incorporated.Then,the affine invariant local regions are built adaptively according to local image content variation.Finally,the watermark is embedded into the local regions by modulating the invariant QPHT modulus coefficients.In this algorithm,the SIFER detector based color invariant theory can improve the repeatability of feature points,in which the important color information is fully employed.Meanwhile a QPHT based color image watermarking scheme is introduced,in which the color image is treated as a vector field and processed directly(with fully considering the correlation between color channels).Experimental results show that the proposed algorithm is not only invisible,but also has a strong ability on resisting common image processing operations and geometric attacks(especially local geometric transformations).3?Based on the theory of Cauchy modeling,we propose an image watermarking algorithm in NSST domain,in which the good balance between robustness and imperceptibility can be well maintained.Firstly,the watermark is embedded into the significant NSST coefficients by using a novel multiplicative method.Then,the watermarked image is geometric corrected by applying the amplitude spectrum of 2D Discrete Fourier Transform(DFT).Next,the Cauchy distribution is utilized to model the NSST coefficients,in which fully employ the sub-band coefficients correlation and RSS method.Finally,the Locally Optimum Detection(LOD)is derived to extract the specific watermark bits based on Cauchy model and Local Most Powerful(LMP)test.In this paper,the Cauchy model is optimized by using the Ranked Set Sampled(RSS)method and the correlation of sub-band coefficients.Meanwhile,the LOD detector is developed using the excellent Cauchy model,which can improve the accuracy of watermark detection.The comparison results show that the proposed algorithm is superior to other algorithms in terms of imperceptibility and robustness.