A Technique to Reconstruct Wavelet-Based Watermark Immune Against Salt & Pepper Noise

This paper proposes a technique to reconstruct watermark immune against salt & pepper noise attack. In watermarking technique, a particular cover embeds an original image as a watermark. To transform the cover image, it uses DWT that gains the LL sub-band, and later, transforms by DCT. To generate and embed the watermark, it modifies the transformed coefficients and the cover image by varying scaling factors. Now to obtain the watermarked image, it applies inverse discrete cosine with DWT on the reformed sub-bands. Next, to extract the watermark, an extraction process is needed. Besides, if high-density salt & pepper noise is enforced on a watermarked image, it exploits the ‘switching threshold algorithm’ for noise elimination, that is implemented in two steps. At first, it detects corrupted pixels. Then using an adaptive length window, it projects new pixel-values only for corrupted pixels. Finally, to validate the effectiveness, the developed technique is assessed.