A new DCT based robust image watermarking scheme using cellular automata

Abstract

ABSTRACT Modern research in image watermarking focuses not only on security but also increasing the payload without having any detectable signature within the media. In this paper, a new Discrete Cosine Transform (DCT)-based image watermarking scheme has been proposed using Cellular Automata (CA). Here, a color cover image is split into three color channels: Red, Green and Blue. Then, DCT is applied on 8 x 8 non-overlapping blocks of each channel followed by a ZigZag scanning. Before embedding, the watermark is encrypted by applying CA rule-15 to make it more secure and robust. The encrypted watermark logo is embedded through DCT coefficients modification using CA rule-340 and mapping table. The proposed scheme is compared with existing state-of-the-art algorithms with respect to embedding capacity, imperceptibility and robustness. Results of the simulation show good imperceptibility measured in terms of Peak Signal-to-Noise Ratio (PSNR) with an average value of 54 dB and simultaneously having high payload of 1.48 bpp. Experimental results under different standard attacks show the superiority of the proposed algorithm as compared with other schemes in the field of image watermarking.