A Robust, Non-blind High Capacity & Secure Digital Watermarking Scheme for Image Secret Information, Authentication and Tampering Localization and Recovery via the Discrete Wavelet Transform

Abstract

In today's world accessibility and trade of advance digital information over the web is of key significance. Security and verification of digital information, including digital images, is one of the greatest concerns. For this specific reason, in this paper we propose a non-blind watermarking technique for self-recovery and authentication of hidden images against unauthorized tampering by utilizing the Discrete Wavelet Transform (DWT) and the Arnold's Transform. Due to the nature of the Discrete Wavelet Transform, DWT-based watermarking schemes, in general, have the hidden images size limited to up to ${}^{1}/{}_{4}$ of the carrier image size; that is, for example, for a $256\mathrm{x}56$ -carrier image the size of the hidden images is usually up to $\mathbf{128\mathrm{x}128}$ . In this paper, we show how the proposed DWT-based technique (a) has a moderately high capacity capability as it allows hidden images of the same size as that of the carrier image, (b) uses a self-recovery framework for watermarking and information hiding that allows embedding twice the number of correction matrices as in prior related work. This makes the proposed non-blind technique equipped with about twice as much capability of self-recovery from unauthorized tampering as in prior work, (c) allows unauthorized tampering localization and recovery against unauthorized tampering such as image cropping, blurring, pixel tampering, (d) utilizes all three-color channels of a carrier image for watermarking and information hiding, and (e) adds security in the secret hidden image information. Experimental results are presented to show the potential value of the proposed non-blind method for multiple information hiding of same size and the carrier image, tampering detection/localization, authentication, and tampering self-recovery, including cropping, and robustness of the proposed scheme to JPEG compression and encryption, and other unauthorized tampering.