Graphical Data Steganographic Protection Method Based on Bits Correspondence Scheme
Автор: Zhengbing Hu, Ivan Dychka, Yevgeniya Sulema, Yevhen Radchenko
Журнал: International Journal of Intelligent Systems and Applications(IJISA) @ijisa
Статья в выпуске: 8, 2017 года.
Бесплатный доступ
The proposed method of graphical data protection is a combined crypto-steganographic method. It is based on a bit values transformation according to both a certain Boolean function and a specific scheme of correspondence between MSB and LSB. The scheme of correspondence is considered as a secret key. The proposed method should be used for protection of large amounts of secret graphical data.
Multimedia Data Protection, Steganography
Короткий адрес: https://sciup.org/15010955
IDR: 15010955
Текст научной статьи Graphical Data Steganographic Protection Method Based on Bits Correspondence Scheme
Published Online August 2017 in MECS
public static bool Encrypt (bool a, bool b, bool c) { if (a ^ b == true) { if (c == true)
return b;
else return !b; } else { if (c == true)
return !b;
else return b; } }.
The function Encrypt is called for every pixel of the cover image, which is used for secret data embedding. In every pixel of imgOriginal 12 secret bits are embedded (by four bits per each colour component – R, G, B). An example of bits traversal is shown in Fig. 2.
Let us consider the example. If the cover image has size 1920 x 1080, then the maximal payload capacity of such image in this method is:
1920 ∙ 1080 / 2 = 1036800 bits.
It means that a secret image can consist of 43200 pixels as maximum:
1036800 / 24 = 43200.
In its turn it means that the secret image of size 240 x 180 can be embedded in this cover image but the secret image of size 320 x 240 cannot.
Let both the secret image be converted into the following bits sequence:
hideImBits = 110111101000… and the cover image consist of the following colour values of pixels:
{ (143, 28, 65), (201, 36, 109), (165, 165, 240), (12, 255, 0), …..}.
Thus, the first pixel is represented by the following binary vectors:
( 143, 28, 65 )?10 ,= (10001111,00011100,01000001)(2)
According to the algorithm as well as both the scheme of correspondence between MSBs and LSBs presented in Fig. 1 and the ternary exclusive disjunction truth-value table (Table 1), the following is fulfilled:
B[0] = Encrypt (G[6], B[0], true) = Encrypt (false, true, true) = false
B[1] = Encrypt (R[5], B[1], true) = Encrypt (false, false, true) = true
RGB
|
5 |
2 |
|
3 |
4 |
|
1 |
6 |
Fig.2. An example of bits traversal in a pixel color representation
B[2] = Encrypt (R[7], B[2], false) = Encrypt (true, false, false) = true
B[3] = Encrypt (G[5], B[3], true) = Encrypt (false, false, true) = true, etc.
As the result the following values of the stego-image (cover image with embedded secret bits) are achieved:
(10001001,00011110,01001110) = (137,30,78)
Then the obtained sequence of the graphical data is to be stored in certain graphical file format.
-
III. Parallel Realization
To achieve significant decreasing of time required for data protection procedure, parallel computations can be employed [15].
The basic algorithm has been analysed and it has been implemented for parallel computing. In particular, the procedure of data embedding has been realized as parallelized algorithm.
The outer loop relates to the number of cores. A separate process is created for every core.
The whole bit array of the secret image is divided according to the number of threads. Every thread is devoted to processing of its part of secret data in to the cover image. Thus, all threads use the same cover image, but their work is not overlapped and they have access only to one memory fragment used for storing certain part of data. Threads work with fragments of the secret image bits array by 24 bits (by 8 bits for every colour component – R, G, and B).
Fig.3. The parallel processing of secret graphical data
Since the parallel realization is based on using a PC with multi-core processor but with limited number of cores (up to four), the most reasonable way of the algorithm parallel realization is parallelization of secret data embedding. In this case all processes inside the loop are independent and use own independent variables and counters. It require more recourses for ensuring independence of the process but nevertheless it allows to achieve time efficiency even for 2 cores. Further number of cores is much more then number of cover parallelization of all inner loops is not reasonable, image pixels rows.
because it requires much more resources and at the same The developed parallelized algorithm is presented in time advantage of parallel processing is tangible only if Fig. 3.
Table 2. Methods time efficiency comparison
|
Experiment number |
Cover image size |
Secret image size |
The proposed method |
The method based on data fragmentation |
The method based on 3DES encryption |
The method based on complementary image |
|
1 |
1324×2048 |
22×40 |
6.7 |
26.50 |
26.10 |
26.00 |
|
2 |
82x46 |
25.4 |
33.05 |
29.95 |
32.85 |
|
|
3 |
128×128 |
72.5 |
38.90 |
41.20 |
36.80 |
|
|
4 |
120x180 |
75.65 |
40.8 |
42.2 |
53.85 |
|
|
5 |
140x140 |
81.5 |
41.1 |
39.65 |
58.2 |
|
|
6 |
260x260 |
83.95 |
53.9 |
70.2 |
60.45 |
|
|
7 |
320x213 |
85 |
55 |
71 |
60.95 |
|
|
8 |
400×200 |
86.15 |
60 |
76.6 |
63.55 |
|
|
9 |
432x384 |
89.4 |
85.6 |
131.3 |
77.1 |
|
|
10 |
4096×2048 |
22×40 |
7.70 |
74.60 |
83.20 |
74.00 |
|
11 |
82x46 |
26.25 |
80.8 |
73.3 |
101.3 |
|
|
12 |
128x128 |
69.95 |
94.4 |
90 |
95.1 |
|
|
13 |
120x180 |
71.4 |
95.3 |
92.65 |
101.3 |
|
|
14 |
140x140 |
72.55 |
97.2 |
91.4 |
112.6 |
|
|
15 |
260x260 |
75.75 |
100.2 |
119.15 |
106.25 |
|
|
16 |
320x213 |
77.3 |
102.45 |
116.25 |
105.95 |
|
|
17 |
400x200 |
80.74 |
116 |
127.45 |
112.65 |
|
|
18 |
432×384 |
84.55 |
128.40 |
179.20 |
126.50 |
|
|
19 |
500x500 |
103.55 |
155.55 |
239.3 |
133.7 |
|
|
20 |
800x500 |
129.25 |
200.35 |
337.75 |
156.3 |
|
|
21 |
800x650 |
134.05 |
241.7 |
416.3 |
188.2 |
|
|
22 |
1024×685 |
148.50 |
316.20 |
553.20 |
251.00 |
|
|
23 |
4096x3072 |
22×40 |
7.3 |
118.55 |
23.5 |
152.4 |
|
24 |
82x46 |
34.25 |
133.45 |
124.35 |
142.5 |
|
|
25 |
128x128 |
73.85 |
142.5 |
138.25 |
156.65 |
|
|
26 |
120x180 |
75.6 |
138.95 |
145.7 |
166.85 |
|
|
27 |
140x140 |
79.55 |
141.55 |
151.55 |
160.75 |
|
|
28 |
260x260 |
81.25 |
135.5 |
158.7 |
150.35 |
|
|
29 |
320x213 |
83.95 |
137.9 |
158.4 |
145.4 |
|
|
30 |
400x200 |
84.2 |
138.7 |
161.75 |
157.4 |
|
|
31 |
432×384 |
87.6 |
167.35 |
220.25 |
163.8 |
|
|
32 |
500x500 |
97.75 |
197.9 |
274 |
176.25 |
|
|
33 |
800x500 |
140.05 |
247.5 |
386.45 |
209.1 |
|
|
34 |
800x650 |
146.35 |
284.75 |
455.1 |
231.5 |
|
|
35 |
800x800 |
181.3 |
321.7 |
540.2 |
253.6 |
|
|
36 |
4096x4096 |
22×40 |
6.75 |
136.8 |
127.9 |
163.9 |
|
37 |
82x46 |
31.2 |
147 |
139.7 |
170.1 |
|
|
38 |
128x128 |
70.45 |
169.3 |
173.95 |
157.45 |
|
|
39 |
120x180 |
68.85 |
165.55 |
155.6 |
163 |
|
|
40 |
140x140 |
69.05 |
152.25 |
152.15 |
157.65 |
|
|
41 |
260x260 |
74.75 |
144.35 |
161.15 |
153.4 |
|
|
42 |
320x213 |
77.55 |
145.55 |
163.75 |
155.85 |
|
|
43 |
400x200 |
79.8 |
150.9 |
171.2 |
155.95 |
|
|
44 |
432×384 |
91.75 |
176 |
233.3 |
175.2 |
|
|
45 |
500x500 |
107.15 |
206.85 |
306 |
188.3 |
|
|
46 |
800x500 |
124.3 |
268.2 |
396.85 |
220.55 |
|
|
47 |
800x650 |
144.15 |
295.4 |
467.5 |
241.45 |
|
|
48 |
800x800 |
155.25 |
335.05 |
552.4 |
253 |
|
|
49 |
1024×685 |
168.65 |
352.2 |
593.3 |
270.65 |
|
|
50 |
1024x800 |
183.6 |
400.8 |
678.8 |
287.55 |
Since a user PC can use not only multi-core but also one-core processor, the developed software enables two modes of secret data processing procedure: parallelized and without parallel computations. The software allows automatic selection of better option. The selection is based on analysis of both the secret image size and the processor characteristics.
-
IV. Results Discussion
In order to test the proposed method, the software package has been developed. The software package allows to measure and compare time efficiency of the proposed method and the following methods:
-
1. The method based on data fragmentation.
-
2. The method based on complementary image.
-
3. The method based on 3DES encryption.
The data fragmentation method [16] uses a separable secret key that consists of 2 sub-keys: the Key of Lengths (KL) and the Key of Addresses (KA). The secret graphical data is transformed into one data sequence. This sequence is divided into fragments of a random length defined by the KL. Every fragment is embedded into the cover image by modifying its LSBs. The place of the embedding is specified by a random address according to the KA.
The complementary image method [17, 18] is based on the complementary transformation of the secret data. The complementary transformation consists in the replacement of every byte of the secret data by a byte kept in the cell of the key table. This cell has coordinates equal to the current byte of the secret data (used as the row number) and the current byte of the cover image (used as the column number). The obtained transformed secret data (called the complementary image) is to be embedded into the cover image.
The method based on 3DES encryption includes two main procedures: the encryption of secret data according to DES algorithm [19] and the embedding this encrypted secret data into the cover image.
The series of experiments has been fulfilled, where different combinations of small, medium, and large cover and secret images were used. In Table 2 results of 50 experiments are presented.
As we can see the proposed method allows to achieve the increase of time efficiency in 4-9 times comparatively to other considered methods when a cover image is large. However, the method has similar or worth time efficiency on small cover images.
-
V. Conclusion
The proposed method of graphical data protection is a combined crypto-steganographic method. It is based on a bit values transformation according to a certain Boolean function and a specific scheme of correspondence between MSBs and LSBs. The scheme of correspondence is considered as a secret key.
The Boolean function can be considered as an additional secret key [20, 21]; however, in this research the ternary exclusive disjunction is used.
Since time efficiency is one of important characteristics of steganographic protection methods [15, 22] along with both robustness against attacks and payload capacity, the proposed method has been realized as a parallelized algorithm. It allowed to achieve significant increase of time efficiency (in 4-9 times) comparing with existing crypto-steganographic methods.
However, this increase can be achieved if a large cover image is used. Thus, the conclusion is that the proposed method should be used for protection of large amounts of secret data.
The further development of the proposed method can be application of its basic principle to other types of multimedia data (audio and video).
Список литературы Graphical Data Steganographic Protection Method Based on Bits Correspondence Scheme
- Internet and cloud services – statistics on the use by individuals, http://ec.europa.eu/eurostat/statistics-explained/index.php/Internet_and_ cloud_services_-_statistics_on_the_use_by_individuals.
- T. Morkel, J.H.P. Eloff, M.S. Olivier, An Overview of Image Steganography, Proceedings of the ISSA 2005 New Knowledge Today Conference, 2005, Pretoria, South Africa.
- P. S. L. M. Barreto, H. Y. Kim, and V. Rijmen. Toward a secure public-key blockwise fragile authentication watermarking. In ICIP 2001, pages 494–497, Thessaloniki, Greece, October 2001.
- B. Chen and G. W. Wornell. Quantization index modulation: A class of provably good methods for digital watermarking and information embedding. IEEE Trans. on Information Theory., 47:1423–1443, May 2001.
- J. J. Eggers, R. Bäuml, and B. Girod. A communications approach to image steganography. In Proceedings of SPIE: Electronic Imaging 2002, Security and Watermarking of Multimedia Contents IV, volume 4675, pages 26–37, San Jose, CA, USA, January 2002.
- P. Guillon, T. Furon, and P. Duhamel. Applied public-key steganography. In Proceedings of SPIE 2002, volume 4675, San Jose, CA, USA, January 2002.
- A. Cheddad, J. Condell, K. Curran, P. Mc Kevitt, Digital Image Steganography: Survey and Analysis of Current Methods, in Signal Processing, 2010, Volume 90, Issue 3, pp. 727-752.
- S. Bhattacharyya, I. Banerjee, G. Sanyal, A Survey of Steganography and Steganalysis Technique in Image, Text, Audio and Video as Cover Carrier, in Journal of Global Research in Computer Science, 2011, Volume 2, No. 4, pp. 1-16.
- C. C. Chang, T. S. Chen, L. Z. Chung, A steganographic method based upon JPEG and quantization table modification, Information Sciences 141(1–2)(2002)123–138.
- A. M. Fard, M. Akbarzadeh-T, F. Varasteh-A, A new genetic algorithm approach for secure JPEG steganography, in:Proceedings of IEEE International Conference on Engineering of Intelligent Systems, 22–23 April 2006, pp.1–6.
- A .I. Hashad, A. S. Madani, A. E. M. A. Wahdan, A robust steganography technique using discrete cosine transforminsertion, in:Proceedings of IEEE / ITI Third International Conference on Information and Communications Technology, Enabling Technologies for the New Knowledge Society, 5–6 December 2005, pp.255–264.
- S. K. Bandyopadhyay, I. K. Maitra, An Application of Palette Based Steganography, in International Journal of Computer Applications, Volume 6, No.4, 2010, pp. 24-27.
- Danny Pascale, A Review of RGB Color Spaces ...from xyY to R'G'B', 2003, BabelColor, Canada.
- N. A. Ibraheem, M. M. Hasan, R. Z. Khan, P. K. Mishra, Understanding Color Models: A Review, in ARPN Journal of Science and Technology, Vol. 2, No. 3, 2012, pp. 265-275.
- Dychka, I. A.; Shyrochyn, S. S.; Sulema, Ye. S. Analysis of Parallel Computations Efficiency for User's Private Multimedia Data Protection in Clouds // Naukovi visti NTUU-KPI, 2016, Issue 1, p. 40-46.
- Sulema E.S., Shyrochyn S.S., "Method of image steganography fragmentation and division stehodanyh private key", Scientific and Technical Collection "Legal, regulatory and metrological support of information security in Ukraine", 2012, Issue 1 (22), P. 64-68.
- Sulema E.S., Shyrochyn S.S., "Image steganography method based on complementary image", Journal "Information Security". 2013, Issue 4, p. 345-353.
- Sulema E.S., Shyrochyn S.S., "Steganographic method of protecting data in audio files from complementary image", Vestnik NTUU "KPI". Informatics, Management and Computer Science: Coll. Science. pr. 2014. Vol. 61, P. 85-92.
- Hummert, K. "The PPP Triple-DES Encryption Protocol (3DESE)", 1998.
- Yevgeniya Sulema, "Image Protection Method Based on Binary Operations", in Proceedings of the 23rd IEEE International Conference on Systems, Signals and Image Processing IWSSIP2016, Bratislava, Slovakia, 2016, pp. 295-298.
- Sulema Yevgeniya, Radchenko Yevhen, Method of graphical data steganographic protection based on bits difference transformation // Proceedings of the 8th Scientific Conference for Students and Postgraduates "Applied mathematics and computing (PMK-2016), 2016, p. 254-258.
- M. Kharrazi, H. T. Sencar, N. Memon, Performance study of common image steganography and steganalysis techniques, in Journal of Electronic Imaging, 2006, No. 15 (4), pp. 1-16.
