To deliver more information in coverless information hiding

Coverless information hiding realizes the hiding of secret messages without modifying the carrier, so being able to resist steganalysis algorithms. Many existing hiding methods, however, still face the dilemma of low hiding capacity (i.e., the volume of hidden messages by a single carrier is bounded), which places some restrictions on the application of the methods. To alleviate this problem, in this paper, we explore an effective coverless information hiding method that can deliver more messages through a cover image. A key step of our method is to construct a visual dictionary based on the bag-of-words model. Utilizing the visual dictionary, one participant can generate a binary coding matrix corresponding to the cover image, and then convert the secret messages, based on the matrix, into a coordinate sequence. To enhance the security, the coordinate sequence is encrypted and the ciphertext is transmitted to another participant together with the cover image. Some experiments are implemented to verify the effectiveness of our method, the results and analysis show that our method performs well in terms of undetectability, hiding capacity, and robustness. Compared with other coverless information hiding methods, our method does not need to construct and maintain a large-scale image library and can deliver more messages by only one single image. These characteristics demonstrate that our method has certain application value, and it also broadens the research ideas of coverless information hiding methods.