Eng During last 365 days Approved articles: 1916,   Articles in work: 289 Declined articles: 811 
Library
Articles and journals | Tariffs | Payments | Your profile

Back to contents

Accuracy of Pixel Predication in Background Areas of Digital Images as Part of Stegoanalysis Performed Using the Weighted Stego Method
Bashmakov Daniil Andreevich

post-graduate student of the Department of Computing System Design and Safety at ITMO University (Saint Petersburg National Research University of Information Technologies, Mechanics and Optics)

197101, Russia, Leningradskaya oblast', g. Saint Petersburg, Kronverkskii prospekt, 49

bashmakov.dan@gmail.com

 

 

Abstract.

 The subject of the research is the accuracy of predicting pixels of background areas of static digital images using the Weighted Stego Image method as part of passive resistance to data transmission channels that use the method of embedding the spatial domain of static digital images with a significant share of homogeneous background in the least significant bit. In his research Bashmakov analyzes the dependence of the accuracy of defining the length of an embedded message on the accuracy of pixel prediction of the image background area. The author focuses on an original algorithm Weighted Stego Image and a number of modifications thereof including the AWSPAM version. He analyzes the formula for calculating an embedded message length and the relationship between its accuracy and the accuracy of pixel prediction. The pixel prediction accuracy is viewed as an error distribution between the predicted and actual values. The accuracy of the stegoanalysis algorithm is viewed as the share of false-positive classifications given the share of correct classifications. The author demonstrates that information about surrounding pixels is not enought for an accurate prediction of pixel and proves that there is a certain relationship between the accuracy of pixel predition and accuracy of an embedded message length, or, in other words, accuracy of the classification using the Weighted Stego method. The author also demonstrates that the AWSPAM algorithm can predict pixels with the highest accuracy compared to the original algorithm prediction function and other modifications thereof.

Keywords: steganalysis method accuracy, image spatial domain, statistical steganalysis, passive resistance, least significant bit, binary classification, steganalysis, steganography, steganographic embedding, steganalytic algorithm

DOI:

10.25136/2306-4196.2018.2.25706

Article was received:

12-03-2018


Review date:

19-03-2018


Publish date:

23-04-2018


This article written in Russian. You can find full text of article in Russian here .

References
1.
Steganography: A Powerful Tool for Terrorists and Corporate Spies // Stratfor [Elektronnyi resurs]. Rezhim dostupa: https://www.stratfor.com/analysis/steganography-powerful-tool-terrorists-and-corporate-spies, svobodnyi. Yaz. angl. (data obrashcheniya 22.08.2017).
2.
Gayathri C., Kalpana V. Study on image steganography techniques // International Journal of Engineering and Technology (IJET). 2013. V. 5. P. 572577.
3.
Prokhozhev N., Mikhailichenko O., Sivachev A., Bashmakov D., Korobeynikov A.G. Passive Steganalysis Evaluation: Reliabilities of Modern Quantitative Steganalysis Algorithms // Advances in Intelligent Systems and Computing. 2016. V. 451. R. 8994. doi:10.1007/978-3-319-33816-3_9
4.
Jessica Fridrich, Miroslav Goljan, "On estimation of secret message length in LSB steganography in spatial domain", Proc. SPIE 5306, Security, Steganography, and Watermarking of Multimedia Contents VI, (22 June 2004); doi: 10.1117/12.521350; https://doi.org/10.1117/12.521350
5.
Ker, Andrew. (2007). A Weighted Stego Image Detector for Sequential LSB Replacement. Proceedings-IAS 2007 3rd Internationl Symposium on Information Assurance and Security. 453-456. 10.1109/IAS.2007.71.
6.
Andrew D. Ker, Rainer Böhme, "Revisiting weighted stego-image steganalysis", Proc. SPIE 6819, Security, Forensics, Steganography, and Watermarking of Multimedia Contents X, 681905 (18 March 2008); doi: 10.1117/12.766820; https://doi.org/10.1117/12.766820
7.
Xiaoyi Yu and Noboru Babaguchi, "Weighted stego-Image based steganalysis in multiple least significant bits," 2008 IEEE International Conference on Multimedia and Expo, Hannover, 2008, pp. 265-268.
8.
Bashmakov D.A., Prokhozhev N.N., Mikhailichenko O.V., Sivachev A.V. Primenenie matrits sosedstva pikselei dlya uluchsheniya tochnosti steganoanaliza nepodvizhnykh tsifrovykh izobrazhenii s odnorodnym fonom // Kibernetika i programmirovanie. 2018.- 1.-S.64-72. DOI: 10.25136/2306-4196.2018.1.24919. URL: http://e-notabene.ru/kp/article_24919.html
9.
Bashmakov D.A. Adaptivnoe predskazanie pikselei v gradientnykh oblastyakh dlya uluchsheniya tochnosti steganoanaliza v nepodvizhnykh tsifrovykh izobrazheniyakh // Kibernetika i programmirovanie. 0.- 0.-S.0-0. DOI: 10.25136/2306-4196.0.0.25514. URL: http://e-notabene.ru/kp/article_25514.html (Stat'ya ozhidaet publikatsii)
10.
Prokhozhev N.N., Mikhailichenko O.V., Bashmakov D.A., Sivachev A.V., Korobeinikov A.G. Issledovanie effektivnosti primeneniya statisticheskikh algoritmov kolichestvennogo steganoanaliza v zadache detektirovaniya skrytykh kanalov peredachi informatsii // Programmnye sistemy i vychislitel'nye metody. 2015. 3. S. 281292. doi: 10.7256/2305-6061.2015.3.17233
11.
BOWS2 the 10 000 original images [Elektronnyi resurs]. Rezhim dostupa: http://bows2.ec-lille.fr/, svobodnyi. Yaz. angl. (data obrashcheniya 12.04.2017).
12.
T. Pevny, P. Bas and J. Fridrich, "Steganalysis by Subtractive Pixel Adjacency Matrix," in IEEE Transactions on Information Forensics and Security, vol. 5, no. 2, pp. 215-224, June 2010. doi: 10.1109/TIFS.2010.2045842
13.
Richard Ernest Bellman. Dynamic Programming. Courier Dover Publications, 2003. ISBN 978-0-486-42809-3