Algebraic Decomposition Method Utilized in Optimized Zero Watermarking Technique
Nada Sabeeh Mohammed, Areej M. Abduldaim · 2021
Mathematics is the backbone of most fields of computer science, especially image processing. Linear algebra achieves important results in image processing by the use of algebraic decomposition methods in different trends, the most important of which are the watermarking techniques. The main aim of this paper is to optimize the zero watermarking technique depending on an optimization algorithm and a decomposition method as an algebraic transformation without using any other popular transform such as discrete wavelet transform (DWT), discrete cosine transform (DCT), and lifting wavelet transform LWT. The singular value decomposition (SVD) is regarded as one of the algebraic methods used in this paper to transform the 8*8 blocks of the original grayscale image into the frequency domain to extract the features of the original image. The singular values in the position (1,1) of each diagonal matrix for each block are chosen to generate the feature bits matrix (master-share) to obtain the final zero-secret of the watermark image. The genetic algorithm (GA) is performed on zero-secret to obtain the zero secret sequence that represents the optimal feature bits matrix (master-share) to optimize the zero watermarking technique. The experimental results show that the technique is worked successfully and is robust and resistant against the attacks adopted depending on the test of the robustness and imperceptibility values.