Hybrid domain watermarking approach for authenticated data protection

Abstract

Image watermarking has developed as a prominent research area in the field of data protection. The authenticated data transmitted through the internet is not secure and can be pirated by unauthorized persons. To protect the valid data, Stationary Wavelet Transform (SWT), Discrete Wavelet Transform (DWT) and Singular Value Decomposition (SVD) based image watermarking algorithm is proposed in the hybrid domain. Initially the original and watermark images are decomposed into approximate (A), vertical (V), horizontal (H), and diagonal subbands (D) using SWT. The approximate band (A) is further decomposed into LL and detail (LH, HL, and HH) subbands using DWT. We calculated SVD for LL and HH subbands of original and watermark images to get the singular values. The singular values of the LL and HH subbands are modified to get the watermarked image. The performance of the proposed model is tested on a standard image dataset. The imperceptibility is evaluated using Peak Signal to Noise Ratio (PSNR), Structural Similarity Index (SSIM), and Feature Similarity Index (FSIM) metrics, and the robustness is validated based on the effective extraction of watermarks from the attacked watermarked image in terms of Normalized Cross Correlation (NCC).