A Coverless Image Steganography Method Based on Feature Matrix Mapping

To address the challenge that the current coverless image steganography (CIS) method requires more cover images and more information, a coverless image steganography method based on feature matrix mapping is proposed. In this paper, the low-frequency component of the Discrete Wavelet Transform (DWT) is used as the image eigenvalue, and a robust feature sequence is obtained by comparing the size of the main and minor diagonal blocks with the eigenvalues of other blocks. This paper proposes a feature matrix mapping rule to obtain the mapping matrix of the cover image and the secret information through the feature matrix mapping. The transmission of secret information is achieved through the cover image and the mapping matrix, which solves the problem of large transmission overhead and the need for more cover images and additional information. Experimental results show that this scheme has certain advantages in terms of the number of mapped images and additional information required, and has better security against most image attacks.