Digital Image Watermarking Technique Using Arnold Transform and Lifting

In this paper, we propose a digital image watermarking scheme using Arnold transform. To implement this, a scrambled binary image is embedded into the coefficient of host images via Lifting wavelet transform. Then inverse wavelet transform is applied to get the watermarked image. We have applied to watermark in the transform domain instead of the spatial domain. This is since in a spatial domain the pixels of the cover image is directly manipulated, in contrast with transform domain where the coefficient of pixels is modified after applying different transformation technique that ultimately increases imperceptibility of the scheme. Moreover, nowadays most of the images are compressed in wavelet domain like JPEG 2000. Another important feature of the proposed algorithm is the Arnold scrambling technique which increases the security of the scheme. It is difficult for the attacker to derive the original watermark as well as retrieve the cover image. The proposed digital watermarking system is tested over a large number of standard grayscale (512×512) benchmark images to provide good imperceptibility and robustness. In our case, we have seen that the proposed technique offers the average value of peak signal-to-noise ratio (PSNR), Structural Similarity Index (SSI) before decoding and after decoding are 49.93, 0.97, 51.72 and 0.976721 respectively. The experimental results show that the proposed scheme is robust against different types of attacks.