Performance Analysis of Digital Image Watermarking Techniques using Zernike Moments and Discrete Cosine Transform

Abstract

Digital image watermarking is the technique very widely used for authentication and security of data. Most of the watermarking techniques have been designed to achieve robustness against various geometric and photometric transformations. Research on a number of watermarking techniques has been carried out. Two broad categories of digital image watermarking techniques are: embedding and extracting watermark in spatial domain and other is in frequency domain. In this paper, two watermarking techniques have been discussed and their performances have been analyzed with respect to geometric and photometric transformations. The first technique for embedding and extracting watermark is to modify magnitude of selected Zernike Moments (ZMs) and other is to modify Discrete Cosine (DC) component of blocks of original host image. It has been shown through experimental analysis that ZMs based watermarking technique is rotation invariant and Discrete Cosine Transform (DCT) based watermarking technique is more robust to noise attack. It has also been analyzed that DCT based watermarking technique is more robust to compression as compared to ZMs based watermarking technique.