Research On Coverless Information Hiding Based On Text Retrieval

Information hiding is a technology that embeds information which needs to be protected into public data without being noticed by a third party.Conventional information hiding methods usually embed information by modifying existing covers or generating covers which are similar to the natural carriers.Due to the principle of the conventional information hiding methods,it is generally difficult for them to resist the detection of various steganalysis algorithms,resulting in the covers being detected by a third party and thus destroying the communication.In order to ensure the success rate and security of secret communication,a new information hiding algorithm-coverless information hiding is proposed.The principle of coverless information hiding is different from conventional one.It transmits hidden information by retrieving natural covers with certain features.Coverless information hiding can fundamentally resist steganalysis since the covers used in the transmission are unmodified natural covers.Although coverless information hiding has many advantages,due to the late start of its research,it has several problems,such as the limited language which can be adapted for text coverless information hiding and the relatively low success rate and hiding capacity.This paper makes researchs on the text coverless information hiding methods based on retrieval which uses natural texts as stego-cover,and proposes methods to solve the problems of the existing text coverless information hiding methods.The specific work of this paper is as follows: 1)English Coverless Steganography Based on Binary Features RetrievalThe existing coverless information hiding framework based on tag-keyword retrieval cannot be applied to English texts.In this paper,a feature tag designing method is proposed.This method converts the English texts into binary stream by encoding and use 6 bits as a feature tag to help locate the secret information.Experiments show that the proposed method can be applied to English texts and the hiding capacity of this method is higher than some existing methods.2)Chinese Coverless Information Hiding Method Based on Hybrid Tags RetrievalTo improve the success rate of information hiding,a method using two tag sets is proposed.This method employs hybrid tag sets.At the information hiding phase,if the retrieval of the cover texts fails,this method uses the other tag sets to retrieve again in order to improve the possibility of information hiding.After the hiding phase,the tag used in every cover text is encoded and transmitted to the receiver along with all cover texts by employing the coverless information hiding method to ensure that only the receiver knows the exact tag used in each cover texts.Experiments and analysis show that the proposed method can improve the success rate and capacity of hiding as well as the transmission security.