Watermarking algorithm for authentication and self-recovery of tampered images using DWT

Abstract

Nowadays the use of digital images is a common practice worldwide, they are used for social, commercial or legal purposes among others; therefore it is important to detect malicious alterations in this type of content, to avoid possible social, legal or personal damages. Currently there are various proposed techniques to solve this problem. However, most of these proposals can only protect a region of interest within digital image, making feasible even with these methods, the alteration of certain regions that are not protected. Therefore, this paper proposes a digital image watermarking algorithm, which consists of two stages; the first one is to protect the digital image that is performed by Daubechies DWT, halftoning and QIM methods; the second one is used for authentication, detection and self-recovery of tampered regions, through IDWT, inverse halftoning and median filtering. Experimental results have demonstrated the effectiveness of the proposed algorithm, obtaining average PSNR values of 35 dB in the self-recovery image, even if the watermarked image was subjected to a JPEG compression using a quality factor QF=75.