Comparative Analysis of Digital Image Watermarking Based on Lifting Wavelet Transform and Singular Value Decomposition

Abstract

The various kinds of unauthorized use and modification of data has been constantly increasing day by day and very common due to the enormous growth of multimedia information on the Internet. Thus, authenticity of the data is one of the major issues in the defence data transmission as well as in secured transmission. To overcome this sort of problem, digital watermarking is supposed to be one of the legitimate solutions as it provides possible validation, safe, and secure transmission of secret data. Here, a robust and secure digital image watermarking method for the protection of copyright has been developed which involves LWT in combination with SVD. The latest approach to wavelet transform is LWT while the significant transform technique is SVD and the obtained experimental results are cross-validated by using inverse LWT and SVD. In order to create and insert a digital signature, the signature mechanism has been used after embedding the watermark and the ownership is authenticated before extracting watermarks in this research work. The robustness of the proposed method for several common image processing attacks like Mean, Median, Gaussian, Shear, Rotation, and Crop is checked by computing the two significant performance metrics, namely, PSNR and MSE. Here, the similarity is also checked by determining the NCC values. Finally, the watermarking scheme suggested in this work is compared with various existing research works which show that the algorithm used in the proposed scheme works fairly in terms of PSNR, MSE, and NCC.