Robust Image watermarking using Tikhonov regularized image reconstruction technique

A robust watermarking scheme is proposed in this paper. The proposed scheme uses discrete wavelet transformations in conjunction with Hessenberg decomposition and singular value decomposition to decompose the cover image and watermark image. Every single value of the watermark is instantly integrated, utilising the optimal scaling factor, into the solitary component of the cover image after the application of singular value decomposition. To make the attacked watermark image attack-free and high-quality, a robust image reconstruction technique is integrated with the proposed watermarking scheme, considering various image processing and geometrical attack scenarios. The proposed watermarking scheme is more resistant to noise-based attacks, according to experimental findings computed in terms of several performance metrics, including peak signal-to-noise ratio (PSNR), structure similarity index (SSIM), bit error rate (BER), and normalised absolute error (NAE).