Embedding and non-blind extraction of watermark data in images in FRFT domain

Abstract

In this paper, a novel scheme to extract the watermark data embedded in FRFT domain is presented. The watermark data (FRFT coefficients of the watermark image) is embedded by adding in FRFT coefficients of a host image at pre-determined locations. A non-blind extraction scheme is the main contribution of this paper. To the best of our knowledge, most of the existing work deals with the watermark detection in FRFT domain and only limited efforts are made for extraction of image watermarking in FRFT domain. But, in this paper, extraction of the watermark was performed in images using original host image. The watermark is extracted by subtracting the FRFT coefficients of the watermarked image from the FRFT coefficients of the original host image. Then the watermark is obtained by performing inverse FRFT of difference FRFT coefficients. Simulations results show that the extracted watermark image is very close to the original watermark image.