Reversible Blind Image Watermarking based on Integer Wavelet Transform

الملخص

In this paper a new reversible (also known as invertible) blind image watermarking technique is proposed based on Integer Wavelet Transform (IWT). This lossless image watermarking technique is developed for some applications like face recognition that a high perceptual quality image is required in order to preserve the performance. For this aim, IWT is applied on an image to decompose it to different sub-bands. Depending on the desired robustness of the watermark, the wavelet’s sub-band is selected. Thus, a lossless function is used to embed the watermark into pair of wavelet’s values. Finally watermarked image is reconstructed by applying inverse IWT. At receiver side, watermark data is extracted from the watermarked image. Then, the original image is recovered by using extracted watermark. Experimental results reveal that the proposed watermark technique not only has proper invisibility but also it has high payload.