Digital watermarking in neural networks models

Digital Watermarking is the act of hiding a message related to digital signal (i.e an image, song, and video) within the signal itself. Watermarking tries to hide a message related to the actual content of the digital image. This paper present two digital watermarking techniques for embedding a text watermark image into gray scale image. This proposed method uses FCNN and Hopfield Model for embed the watermark to achieve almost zero visible distortion in the watermarked image. Therefore watermarked image is almost same as the original cover image and extracted watermark at the output is same as the watermark at the input. Performance of these two models is compared on the basis of nop, Elapsed time, PSNR before adding a watermark, PSNR for extracted watermark image, Attack recover time, Ncor.