Efficient Robust Reversible Watermarking Based on ZMs and Integer Wavelet Transform

Abstract

The robustness of existing robust reversible watermarking (RRW) against attacks cannot satisfy requirement for image copyright protection of industrial Internet of Things. To address these issues, in this article, we propose an efficient robust reversible watermarking scheme based on ZMs and integer wavelet transform for industrial images. First, Zernike moments (ZMs) in the low-frequency region of integer wavelet transform coefficients of the original image are calculated. Then, the multibit watermarking is reversibly embedded into the magnitudes of the ZMs. The errors generated when embedding the watermark are reversibly embedded into the pixels of the image. Later, the receiver can utilize the errors extracted from the watermarked image to restore the original image. In the case of images being attacked, the extracted robust watermark can be used as a proof of copyright authentication.The experimental results demonstrate that the proposed method can not only completely restore the original industrial image when the watermarked image is not attacked, but also has stronger robustness against geometric transformations, common image processing operations and joint attacks compared with the state-of-the-art RRW authentication schemes.