Image Watermarking Technology Based On Hypercomplex Continuous Orthogonal Moments

The Internet’s rapid development has led to widespread data sharing and openness,which has also given rise to increasingly serious infringement problems.Ensuring the safety of digital resources during storage and dissemination has become a widely concerning issue in society.Digital watermarking has emerged as a solution to image copyright protection and has quickly become a hot topic in research.Research on image watermarking has evolved from grayscale images to color images and more recently to stereo image watermarking technology.However,for multi-view color images and light-field images with complex structures,there are many problems with watermarking technology in terms of resistance to geometric attacks and visual quality,and there is no mature or perfect solution yet.Image moments have good global feature description ability and geometric invariance,and their application to image watermarking can effectively solve the problem of anti-geometric attack.Although existing hypercomplex continuous orthogonal moments have allowed for the rapid development of color and binocular stereo image watermarking,they still cannot handle multi-view color images and light-field images with more complex structures.To solve these problems,three new types of hypercomplex continuous orthogonal moments are proposed in this thesis,on top of which four watermarking methods are proposed.The main work of this thesis is as follows:(1)Fractional-order Chebyshev-Fourier moments(Fr CHFMs)are constructed based on the Chebyshev-Fourier moments(CHFMs),which solve the numerical instability of the radial basis function of CHFMs effectively.On this basis,combined with the sedenion theory,sedenion fractional-order Chebyshev-Fourier moments(SFr CHFMs)are constructed.The application of the hypercomplex moments is extended from color images and stereo images to multi-view color images.Using SFr CHFMs for constructing zero-watermarking of multi-view color images enables lossless copyright protection of multi-view color images and shows good robustness against various attacks.(2)Relying on geranion theory and polar harmonic Fourier moments(PHFMs),geranion polar harmonic Fourier moments(GPHFMs)are constructed and utilized for light-field image watermarking.The geranion represents a hypercomplex number containing one real part and thirty-one imaginary parts,and the imaginary parts of geranion can be used to encode multiple image color components while maintaining the correlation between each component.GPHFMs are stable image features,with good image reconstruction ability and stability.Essentially,the light-field image watermarking based on GPHFMs offers good imperceptibility and robustness,can resist various attacks,and effectively solves the problem of current watermarking schemes related to their inability of resisting geometric attacks.(3)A multiple reversible robust light-field image watermarking scheme based on code division multiplexing(CDM)and quaternion polar harmonic Fourier moments(QPHFMs)is proposed.The scheme which embeds multiple watermarks into the light-field macro-pixel image and the information loss caused by embedding multiple watermarks is reversibly embedded into the light-field sub-aperture images.The geometric invariance of QPHFMs and the fault tolerance of multiple watermarks significantly improve the watermarking scheme’s resistance to geometric attacks.The CDM-based reversible information hiding is zero mean,which ensures that the light-field image still has high visual quality even after multi-level information embedding.The proposed watermarking scheme can extract watermark information and fully recover the original light-field image when the image has not been attacked,and can also extract robust watermark information for verifying the copyright ownership of the light-field image when the image has been attacked.(4)A generalized hypercomplex continuous orthogonal moments(HCOMs)are proposed and extended to multi-dimensional hypercomplex continuous orthogonal moments(MHCOMs)in multi-dimensional space.MHCOMs can be used to process images with arbitrary multi-dimensional structures,and when they are used for light-field image processing,there is no need to map the light-field images to two-dimensional space to change the multi-dimensional structural properties of light-field images.By calculating MHCOMs of light-field images,a set of multi-dimensional hypercomplex tensors containing robust feature information of light-field images can be obtained,and the original light-field images can be reconstructed using this information.Furthermore,the MHCOMs-based zero-watermarking scheme is designed for light-field images,which has excellent robustness and can achieve lossless copyright protection for light-field images.