Scientific publications
Damaris Belle M. Fotso
Other affiliations: Unité de Modélisation Mathématique et Informatique des Systèmes Complexes · Institut de Recherche pour le Développement
Stéphane Gael Raymond Ekodeck
Other affiliations: Unité de Modélisation Mathématique et Informatique des Systèmes Complexes · Institut de Recherche pour le Développement
Other affiliations: Unité de Modélisation Mathématique et Informatique des Systèmes Complexes · Institut de Recherche pour le Développement
Research Square
Abstract Rather than directly embed the secret message in pixels of an image, the new concept of coverless image steganography hides a message in a cover image without modifying it, which significantly increases the resilience to image steganalysis tools. Existing high capacity coverless image steganography methods do not provide an efficient algorithm for rapidly generating the large number of possible hash sequences. Moreover, the sending of additional information generated during the hiding process and essential for extraction is not explicitly defined. In this paper, we propose WYSAWIS, a high capacity coverless image steganography approach. Given an image initially subdivided into blocks of size 4*4, we generate from each block an initial 15-bit hash sequence, then a set of all the combinations of this sequence. The latter constitutes the set of hash sequences of that block.To send a secret message, it is first subdivided into 15-bit segments,and each segment is mapped to a hash sequence of a block, whose position in the block determines the position of the message segment. All these positions called location information are sent using a distributed data hiding technique in a single cloud. Compared to what has been done in the literature before, our hashing algorithm makes it possible to use a single cover image to send a secret message of any size. This definitely overcomes the limits of building a large image data set and sending in order a large number of stego images. Moreover, the popularization of cloud services makes the transfer of location information innocuous and practical
