Image watermarking with path based selection using DWT & SVD

This paper, the robustness of the proposed watermarking algorithm over various attacks (speckle noise, salt & pepper noise, brightness, Gaussian noise, rotation & contrast) is verified. In the proposed embedding, the embedding carries with the wavelet & Singular value decomposition (SVD) transform coefficients of both the images (cover & watermark) & interpolation. Here the SVD coefficients of the watermark is embed in row by row & column by column sequence of the cover image's SVD coefficients. This process of embedding ends more security & protection to the watermarking technique. The inverse of embedding is taken to extract the watermark image. By using Peak Signal to noise ratio (PSNR) & Mean square error (MSE) parameters the comparison is done with the basic DWT [16] & path based DWT [17] for better efficiency verification.