Contrast Media Mol Imaging. 2021 Dec 21;2021:5483001. doi: 10.1155/2021/5483001. eCollection 2021.

Reversible Information of Encrypted Image Based on Feature Difference Detection and Wavelet Transform.

Contrast media & molecular imaging

Yongsheng Ding, Yunbo Wei, Shuisheng Zhang, Shihang Yu

PMID: 35024013 PMCID: PMC8714395 DOI: 10.1155/2021/5483001

Abstract

Aiming at the shortcomings of the existing lossless digital watermarking algorithm based on frequency domain in reversibility and embedding capacity, this study proposes a lossless digital image watermarking algorithm based on fractional wavelet transform, which is used for large-capacity reversible information hiding of images. First, the image is transformed by LeGall5/3 fractional wavelet, and then, the watermark is embedded in the high-frequency subband by the histogram shift method. In order to obtain maximum embedding capacity and reduce image distortion, the methods of selecting embedding parameters and stopping parameters are proposed, respectively. At the same time, in order to prevent overflow and reduce additional information, a new method of generating position map is proposed. The experimental results show that Lena is the result of multilayer embedding based on the algorithm in this study. In order to better observe the distortion phenomenon and enlarge the image, the Lena test image is the watermark image obtained after two and three layers of embedding, and its embedding capacity can be 2.7 bpp. It is proved that wavelet transform is suitable for encrypted images to implement covert communication.

Copyright © 2021 Yongsheng Ding et al.

Conflict of interest statement

Publication Types

• Journal Article