Watermarking Framework for Authentication and Self-recovery of Tampered Colour Images

Methods of authentication and self-recovery of tampered information in digital images have been in constant development during the last years. These frameworks are developed by watermarking techniques, which generate an image digest, which is embedded inside the image that must be protected. In this paper, a method of authentication and self-recovery of tampered information in digital images is proposed. Our approach first generates the authentication watermarks, which are based on XOR operations on non-overlapping blocks, subsequently by using a halftoning technique the recovery watermark is generated. To have a higher performance, three copies of each watermark are embedded to have more chance to extract the recovery watermark. To evaluate the quality of the obtained images, the objective criterion of peak signal-to-noise ratio (PSNR) and Structural Similarity Index (SSIM) are used. The experimental results demonstrate the effectiveness of our method in comparisons with other schemes reported in the literature, where the quality of the watermarked images, the quality of the reconstruction images and the recovery rate of each scheme were evaluated.