A robust chaotic algorithm for digital image steganography

•We model a steganographic frequency-domain algorithm for hiding information in an image.•This model uses DWT coefficients of a cover image according to Sweldens’ lifting scheme.•A 3-D chaotic map is used to select in a pseudorandom manner the host coefficients.•Experimental results show that this algorithm has strong imperceptibility and security.•Advantages over existing frequency-domain steganographic algorithms are highlighted.

This paper proposes a new robust chaotic algorithm for digital image steganography based on a 3-dimensional chaotic cat map and lifted discrete wavelet transforms. The irregular outputs of the cat map are used to embed a secret message in a digital cover image. Discrete wavelet transforms are used to provide robustness. Sweldens’ lifting scheme is applied to ensure integer-to-integer transforms, thus improving the robustness of the algorithm. The suggested scheme is fast, efficient and flexible. Empirical results are presented to showcase the satisfactory performance of our proposed steganographic scheme in terms of its effectiveness (imperceptibility and security) and feasibility. Comparison with some existing transform domain steganographic schemes is also presented.