Image Processing and steganography Research Papers (original) (raw)

Abstract Steganography is a “science”, the method of hiding sent information. It embeds the secret message in cover media (image, audio, video, text, etc.). The most popular steganography method is LSB (Last Significant Bit) replacement... more

Abstract
Steganography is a “science”, the method of hiding sent information. It embeds the secret message in cover media (image, audio, video, text, etc.). The most popular steganography method is LSB (Last Significant Bit) replacement in the cover image. The most notable Steganalysis algorithm is the RS method [1], which detects stegamesage by the statistical analysis applied on image pixels. The goal of the project is to demonstrate effectiveness of improved GSM algorithm proposed by Shen Wang and others [2] against RS analysis.
Key words: Steganography, Steganalysis, LSB, Digital image.
1. Introduction and problem definition
In modern word information is has great value. With global computer networks appearance volume of transmitted and received information has been increased, a lot of data transferred via global webs. And as results of easy accessibility to different information, sometimes to high sensitive information, there is a need to protect data security and threat unauthorized access to information. On other hand, with advancements in digital communication technology and the growth of computer power and storage, the difficulties in ensuring individuals’ privacy become increasingly challenging. Data, intellectual property and privacy protection – this is scabrous problem with that we face on a daily basis. Various methods have been investigated and developed to perform data protection and personal privacy. Encryption is probably the most obvious one, and then comes steganography. Steganography is the art and science of invisible communication. This is accomplished through hiding information in other information, thus hiding the existence of the communicated information. The word steganography is derived from the Greek words “stegos” meaning “cover” and “grafia” meaning “writing” defining it as “covered writing”. In general, steganography approaches hide a message in a cover e.g. text, image, audio file, etc., in such a way that is assumed to look innocent and there for would not raise suspicion [3]. Except to transfer secret information or embed secret messages into media, one of important and perspective application of steganography is to protect intellectual property and copyright on digital media, images, books to avoid unauthorized copying and theft. The main purpose of this work is to study LSB based Steganographic and Steganalysis methods.
Implement and study RS Fridrich algorithm [1]. In second part of work introduce modified “Genetic shifting Algorithm” proposed by Shen Wang [2], method of embedding secret message in to digital image, without causing visual degradation of cover/stego image and to avoid stegamesage presence detection by RS Analysis algorithm.
Opposite to Shen Wang steganography method, which performs final stegoimage bits manipulation, this paper is deals with original “cover” image. Changes are made in cover image with target to “worsen” bits statistics. And as a result of this permutations, secret message embedding provides “positive” statistics changes that affect RS analysis determine message existence.
2. Project steps
The current project objectives are:
 Perform comparison visual and statistical analysis for different message length.
 Check what message length can be embedded into cover image without visual or statistical image degradation.
 Check dependence of the image degradation from embedded message length.
 New Stego optimized Genetic Shifting Algorithm definition.
 Confirm effectiveness of new method in interaction with Fridrich RS algorithm, for Grey scale images.
All experiments and research have been performed in LabVIEW environment, and include next steps:
1. Build LabVIEW based working steganography model
for embedding text message into digital image with LSB embedding algorithm. According to LSB method, every text and every image can be transferred into binary form, after this operation we can replace least significant bit of image by bits of message. After successful LSB-1 encoding and image recovery, perform same experiment on same images up to LSB-4. LSB-1 and LSB-4 signs level of significant bits permutations, there LSB-1 – only one least bit is used and LSB-4 – four significant bits are used.
2. Perform basic message coding (Cover Image) and
recovery up to LSB-4 for gray images.
Figure 1 Demonstrates message embedding steps,
סיכום פרויקט
Application of improved Steganographic Genetic Shifting algorithm against RS analysis
פורינסון ודים
קפלן ולדיסלב MScEE
Figure 1. Basis diagram for message embedding.
3. Compare visual image degradation.
4. Compare visual degradation through common
tools
(Histogram, STD). With Statistical tools help perform image degradation analysis versus LSB level and imbedded message volume.
5. Perform study of coded message saturation
(message of different length) vs. recovery and image degradation per different LSB coding at gray images.
6. Build RS analysis (Fridrich algorithm) [1]
routine.
7. Confirm validity of RS analysis on gray images.
Figure 2. Demonstrates resulting RS analysis plot.
Figure 2. RS analysis plot
8. Implement secure genetic steganography method
for RS baseline shifting for LSB-1. (GSM for RS shifting).
9. Perform basic message recovery with GSM for
RS shifting for LSB-1.
10. Perform RS analysis comparison for different
message length with GSM for RS shifting and without, use different “snake” division array image representation.
3. Results and discussion
This project is meet all objectives was targeting in start of the work:
 Working LabVIEW based model for secret message to image encoding/decoding is implemented.
 Study images visual degradation by “naked” eye and Compare visual degradation through common tools (Histogram, STD).
 Study Fridrich RS analysis algorithm is performed.
 Checked and proved RS algorithm validity for secret message presence detection into Grey scale images.
 Determined strong dependence of image visual degradation on message length (volume) and depth of LSB levels manipulations.
 Definition and improvement of existing Shen Wang and al [2] Genetic Shifting algorithm.
 Checked and proved ability of proposed Shifting algorithm to against RS attack.
4. Summary and conclusions
In most of the original digital images exists a high matching between the pixels that are placed next to each other [2], in case any bit manipulation is performed this causes a matching between pixels is worsens. This is reason for high histogram sensitivity for any bits replacements. But in same time, we can see, that LSB-1 level do not dramatically impact image histogram, and in case no clean image histogram presents to compare, this is impossible to determinate stegamesage is exists. By using received statistical data we can with high probability determine embedded message existence and approximate message length. Another words, 𝑅𝑚−𝑆𝑚 differences (under normal conditions) less 7% indicates LSB manipulations with high probability. Recess of LSB levels manipulations improve RS analysis stability to determine embedded messages, but in this case visual attack is prefer and easy.
5. Literature
[1] Jessica Fridrich, Miroslav Goljan “Practical Steganalysis of Digital Images – State of the Art.” SUNY Binghamton, Department of Electrical Engineering, Binghamton, NY 13902-6000.
[2] Shen Wang, Bian Yang and Xiamu Niu, “A secure Steganography Method based on Genetic
Algorithm”, School of computer Science and Technology Harbin Institute of Technology.
[3] E.L. Zorin, N.V. Chichvarin “Steganography”, MGTU. N.E. Bauman, Faculty "Information and Management", Department "Information Security", Moscow 2011.
[4] E.V Selantev “Fundamentals of computer steganography”, Moscow Institute of Electronic Engineering, Faculty of Information, Moscow 2009.
[5] C.P.Sumathi, T.Santanam and G.Umamaheswari “A Study of Various Steganographic Techniques Used for Information Hiding” International Journal of Computer Science & Engineering Survey (IJCSES),Vol.4, No.6, December 2013.
[6] Arooj Nissar , A.H. Mir. “Classification of steganalysis techniques”, Department of Information Technology, National Institute of Technology, Srinagar 190006, India.
[7] LabVIEW “Getting Startedwith LabVIEW” National Instruments, June 2009, 373427F-01.
[8]http://bilder.buecher.de/zusatz/22/22359/22359536_lese_1.pdf. “A survey of steganographic techniques” Chapter 3.