Abstract
In this paper, DWT-based image watermarking methods are presented and evaluated with a view to improving their robustness and imperceptibility. In this approach, watermarks are embedded and extracted at various levels of decomposition, including one-level, two-level, and three-level. Simulations are conducted in MATLAB. The algorithm was evaluated using peak signal-to-noise ratios (PSNRs) and normalized cross-correlations (NCCs) across different host and watermark images. PSNR and NCC values increase when the decomposition level increases, demonstrating improved watermarking performance. Graphs and tables were used to compare these two levels of decomposition. Researchers have demonstrated that DWT-based watermarking is capable of providing secure image communication as well as data protection. Computer vision researchers, image processors, and data security experts can benefit from this study by improving watermarking methods.
View more >>