Font Size: a A A

Research Of Embedded Messages Extraction Method On Image Adaptive Steganography

Posted on:2019-06-24Degree:MasterType:Thesis
Country:ChinaCandidate:J J GanFull Text:PDF
GTID:2428330566470910Subject:Military cryptography
Abstract/Summary:PDF Full Text Request
Image steganography is the practice of embedding messages into images,which not only hides the contents of the messages but also the fact that communication is underway.Image steganalysis is an inverse technique of image steganography.It mainly research on the detection and extraction of embedded messages.There are no significant difference in senses and statistics between the cover images and the original images,therefore,image steganalysis is a research direction with strong exploratory and high degree of difficulty.The embedded messages extraction is the ultimate purpose of steganalysis,with great practical significance to obtain covert communication content and covert communication forensics.Image adaptive steganography constrains the embedding modifications to the areas difficult to model,with strong concealment and anti-detection performance.So far,there are few reports on the embedded messages extraction of image adaptive steganography.The embedded messages extraction of image adaptive steganography are realized when the embedded messages are plaintext and other two harsh conditions.The embedded messages extraction method with wide suitability under ciphertext embedding are urgent to be studied.In the first,the security of the STC(Syndrome-Trellis Codes)is studied,and then the embedded messages extraction method of three common application scenarios are studied on this basis,the contributions are as follows:1.The principle of STC is indicated,the security of the STC is studied,research indicates: STC has the vulnerability to algebraic attacks and imperfect secrecy.on the basis,the extraction frame for image adaptive steganography is proposed.Based on the framework,a variety of embedded messages extraction method can be designed for different application scenarios.2.Extraction method for image adaptive steganography based on algebraic attack is proposed when the partially known embedded messages are known.The simplification method of the STC decoding equation is given based on the identical deformation of the matrix operation.the non-homogeneous linear equations about the elements of sub-check matrix are obtained based on the vulnerability to algebraic attacks of STC.The number of known embedded messages bits at different embedding rates are theoretically analyzed.And the experimental results show that the parameter recognition of STC can be achieved when the partially known embedded messages are the name of file format and the payload is more then 1/3,and then the embedded messages extraction for image adaptive steganography is achieved.3.Extraction method for image adaptive steganography based on the existence of solution in STC decoding equations is proposed when the embedded messages are resented.Based on the imperfection of STC and the existence of solution in STC decoding equations,the non-homogeneous linear equations about the embedded messages are obtained,multiple stego images with same embedded messages are utilized to reduce the entropy of embedded messages space.And then,the times of resending are theoretical analysis.Experimental results show that when the relative payload is greater than 1/32 and the same embedded messages are repeatedly transmitted twice or more,then the embedded messages can be correctly extracted.4.Extraction method for image adaptive steganography based on the parametric hypothesis test is proposed when the keys are reused.Using the encryption characteristics of same keys to find the distribution difference between the messages extracted by the true and false coding parameters,a coding parameters recognition method based on parametric hypothesis testing is given.Based on the probabilities of type I errors and type II errors,the formulas of sample size and threshold are deduced.are presented for calculating the for the proposed stego key recovery model.Experimental results show that when the relative payload is greater than 1/3 and the embedding key is repeatedly used for 11 times or more,then the coding parameters can be correctly identified,the embedded messages can be correctly extracted.Finally,the research work of this thesis is summarized and the future research work for extraction research of image adaptive steganography is prospected.
Keywords/Search Tags:Information hiding, Image adaptive steganography, STC, Ciphertext embedding, Extraction attacks, Parameters recognition
PDF Full Text Request
Related items