Robust coverless steganography using limited mapping images

Coverless image steganography (CIS) has attracted significant attention because it can fundamentally resist steganalysis tools. Available CIS schemes are mainly divided into synthesis-based and mappingbased schemes. Compared with the former, mapping-based methods ensure lossless information extraction and stronger attack robustness. However, these methods still face the challenges of incomplete secret information mapping, trade-off between robustness and distinguishability of mapping features, and demand for huge numbers of mapping images. To tackle these issues, a novel robust coverless steganography using limited mapping images is proposed in this paper. In our scheme, we extract ring statistics to ensure both the distinguishability and robustness of mapping features. Moreover, different from conventional CIS schemes, we further design a chaotic system for scrambling image features to map secret information. Using this system, a single image is mapped to multiple secret information instances using different scrambled features, which reduces the demand on the numbers of images and avoids incomplete mapping. Furthermore, the scrambled features also enhance the security of secret information. Experimental results demonstrate that our scheme has superior performance in terms of distinguishability and robustness against geometrical attacks, the required number of mapping images, and mapping completeness. (C) 2022 The Author(s). Published by Elsevier B.V. on behalf of King Saud University.

Automated summary

This paper proposes a novel robust coverless steganography using limited mapping images to address the challenges of incomplete mapping, robustness, and distinguishability. By extracting ring statistics and using a chaotic system to scramble image features for mapping secret information, the proposed scheme reduces the demand for mapping images and enhances information security.