A Watermarking Scheme for Digital Images Developed Using Discrete Cosine Transform (DCT)

In this research, a grayscale image watermarking approach is introduced, utilizing a discrete cosine transform (DCT) to embed an imperceptible binary watermark image. Diverging from traditional methods that partition the image into non-overlapping 8x8 blocks, the DCT is employed across the entire host image. Following this, specific coefficients from the low-frequency DCT range are selected for the embedding of the watermark. In order to attain a harmonious equilibrium between the resilience of the watermark and its inconspicuousness, the intensity of watermarking is controlled through a robustness factor. The precise value of this robustness factor is flexible and can be fine-tuned by the user in accordance with their preferred level of resilience and the attributes of the host image. Assessing the quality of watermarked images involves the computation of peak signal-to-noise ratio (PSNR) and normalized correlation (NC). The conducted simulations and resultant findings highlight the resilience of the proposed methodology, showcasing its efficacy even under a variety of attacks, encompassing JPEG compression, low-pass filtering, and noise distortions