Digital image watermarking using optimized transform-domain approach

Digital watermarking under varying conditions may prove challenging for the safe retrieval of data. This paper puts forward a novel algorithm, which employs a unique combination of Discrete Wavelet Transform (DWT), Discrete Fourier Transform (DFT), Discrete Cosine Transform (DCT) and Singular Value Decomposition (SVD), to provide an effective approach for efficient watermarking. The DWT stage selectively utilizes the horizontal detail coefficients of the 2-dimensional DWT of the image. The translation variance problem of the DWT is rectified in the DFT stage that follows, which uses the frequency characteristics of the image. Then, DCT is used to compress because of its efficient compression representation power. Finally, SVD is used to embed the watermark into the data. The results obtained after applying the proposed algorithm on the standard test images, show that the proposed system outperforms other systems. A comprehensive rise in the PSNR value is observed.