DWT SVD Based Segmented Watermarking Scheme Using Genetic Algorithm

A multiple color image segmented watermarking scheme based on discrete wavelet transform (DWT) and singular value decomposition (SVD) is presented. In the proposed approach, the original image is segmented into two sub images, and then the two color watermarks are embedded in the singular values of each sub image separately. In the extraction process the watermarks are extracted from the singular values of the watermarked sub images. The segmentation of multiple watermarking processes makes the watermarks much more robust to the attacks such as noise, filtering, compression, rotation, cropping, translation, sharpening, smoothing, row-column blanking Intensity transformation, and row-column copying. The optimization on segmented watermarking achieves more imperceptibility and robustness.