- Open Access
- Total Downloads : 2
- Authors : J.Jayaseelan, B.Kruthika
- Paper ID : IJERTCONV2IS05034
- Volume & Issue : NCICCT – 2014 (Volume 2 – Issue 05)
- Published (First Online): 30-07-2018
- ISSN (Online) : 2278-0181
- Publisher Name : IJERT
- License: This work is licensed under a Creative Commons Attribution 4.0 International License
A Novel Noise Guided Random Stegging with Adaptive K-bit Engrafting to Enhance Eminent Concealing Capacity
Co mmunicat ion Systems,
Parisutham Institute of Technology & Science, Affiliated to Anna University Chennai.
Tamil Nadu- India
Ema il: email@example.com
Parisutham Institute of Technology and Science, Affiliated to Anna University Chennai.
Tamil Nadu- India
Abstract- In this paper a novel noise guided clandestine data engrafting in a binding image is proposed. In the proposed method two copies of binding image is carried. A pixel value of one binding image is changed by the addition of noises such as S alt & Pepper, Gaussian, Etc., and represented as noisy image. By reference of this noisy image the data needs to be secured is engrafted into the binding image. Besides protection of data, the quantity of data that can be concealed in a single bearing medium is also very important. This high engrafting capacity is attained by k- bits of clandestine messages are substituted in k- least significant bits of image pixels, where k alters from 1 to 3 depending on the added noise. The proposed scheme is examined and results compared with existing single bit substitution for the test images temple, Gandhi, baboon and Lena. The experiment results affirm that the proposed scheme attains eminent data concealing capacity and maintains imperceptibility and dilutes the aberration among binding image and obtained stego image
Index terms: Binding image, clandestine data, k- bit embedding, Noise guided stegging, stego image
The shelter of confidential data has long been a major pertain. To defend this e xtre me ly confidentia l data fro m being tapped, altered, or utilized by unauthorized persons, we required having methods for attaining data protection.
The most well known method for data protection is using Data Concealing, which is the procedure of concealing details of an article or function. A significant method of data concealing is steganography [1- 5]. It is the skill of concealing data. It conceals the clandestine message within the emcee data set and makes its presence imperceptible. The ma in objective of steganography is to avoid absorbing hunch to the existence of a concealed content. [6-9]
In steganography, the binding mediu m is the file in which we will hide the clandestine data. The resultant file is the stego mediu m. The binding med iu ms are typically image or audio files. [1, 2, 10-14]
RELAT ED WORKS
In the Recent days, lots of steganography methods have been suggested. They are separated into two classification accomplished on their binding image do ma ins: videlicet, spatial and frequency [1, 2, 15-19]. In the spatial domain, the secret entropy are concealed in the pixe ls of the binding image by applying Least Significant Bit (LSB) [20, 21], Pixe l Value Diffe rencing (PVD) , mod, run-length reversible and lossless informat ion concealing based strategies. These strategies have been employed by many researchers to achieve beneficial impe rceptibility with a more e minent consignment. In the frequency domain methods, the clandestine informations are concealed in the transformed coeffic ients of the binding image, where Discrete Cosine Transform (DCT) and Discrete Wavelet Transform (DWT) play the doma in converters [18, 19, and 27]. Of the spatial domain stego methods, the LSB engrafting strategy has been broadly used to conceal clandestine information because of its simp lic ity and hasten of effectuation, which e xtends a more eminent concealment capability 
To improve the concealing capacity, more nu mber of clandestine data should be engrafted into all binding image pixe ls. Regrettably this scheme abbreviates the lineament of the consequent stego image. Besides the lineament, quantity of information that can be engrafted into an individual binding med iu m is also very significant [23-24, 26].
In our proposed scheme, an adaptive k-bit engrafting technique is employed. It meliorates the concealing capacity without conciliatory the quality of the consequence image. In an existing once the cyberpunks hacked the stego mediu m then the chance of capturing the secret information is eminent. But in this proposed scheme it is insufferable; because each pixe l in binding image is engrafted with dissimila r nu mber of p ixe ls.
In an e xisting LSB substitution techniques preprocessing is done by dividing the binding image into blocks, dividing the binding image into color planes, adopting pixe l indicator
based substitution, Z- scanning, random wa lk methodology etc,. In our paper, binding image of proposed system is preprocessed by the addition of Salt & Pepper Noise. Noise with defined density is added with the copy of binding image.
This added noise alters the binding image pixel va lues to either ze ro if it is added with Salt or Ma ximu m Intensity if it is added with pepper. Finally this noisy image is represented as guiding image.
In this paper, Spatial domain steganography is adopted by employing a Noise guided random stegging with adaptive K- b it engrafting for accomp lishing eminent concealing capacity without conciliatory the caliber of stego image.
A. Noise guided stegging
Salt & Pepper noise is a random noise with ON & OFF Pixe ls. It modifies the pixe l values into either Ze ro or Maximu m intensity of the image. In this proposed scheme Salt & pepper noise is employed for the preprocessing of input binding image. Mostly preprocessin g is done for picking out the pixe l e mplace ments of binding image to engraft the clandestine data. If it acco mpanies any order then the possibility of hacking the secret data is eminent. By the Noise Guided Stegging technique Salt & pepper Noise with determined density is contributed with the input binding image.
This noise added binding image is represented as guiding image or reference image with three dissimilar set of pixe ls they are Salty pixe ls, peppery pixels and pure pixels. Corresponding pixels assesses are Zero, Ma ximu m (255 for
256 * 256 images) and similar as like b inding image respectively. Instead of fixed decision ma king for the number of pixels to be engrafted into a binding image, our proposed scheme avails the user defined decision making p otentiality. A sender or engrafting authority can decide the pixe ls should be utilized for engra fting. Sender has the following choices for concealing the clandestine data: select salty pixe ls of
pixe ls. Now the sender can engraft the clandestine data with the following bit sets as respective of above order:
1-2-3, 1-3-2, 2-1-3, 2-3-1, 3-1-2, and 3-2-1. Th is ability
attains the eminent concealing capacity without conciliatory the quality of stego image. This decision ma king also made as user defined.
System design comprises two contributions s uch as engrafting and retrieving as shown in fig.1. In an engrafting part, encrypted secret data and binding image are afforded to the stego system encoder as inputs. Stego system encoder adopts our proposed system for the process of engrafting the secret data into the binding image with the support of guiding image. In a retriev ing part the reverse process of above is done for acquiring the transmitted clandestine data.
Encrypted Clandestine Data
Stego System Encoder
Decrypted Clandestine Data
Stego System Decoder
Fig.1: System Design
binding image, peppery pixe ls of binding image, pure pixe ls of binding image, both salt & pepper, both salt & pure pixels, Pepper & Pure p ixe ls. So the pixe ls with engrafted clandestine data are extre mely insufferable to accumulate because of the randomizat ion of proposed methodology. Eventually this noise guided random stegging system me liorates the quality of the resultant stego image with high imperceptibility factor.
B. Adaptive K- bit Engrafting
Binding Image (A i, j)
Reference Image (B i, j)
Com pare Pixels of (A i, j) & (B i, j)
(B i, j) = 255
(B i, j) 255 or 0 (B i, j) = 0
Engrafted data set2
Engrafted data set3
As mentioned before the quantity of clandestine data that can be engrafted into a single binding image without fle xib le the linea ment of stego image is very significant. To attain this, adaptive K- bit engrafting technique is proposed. Here K a lters fro m 1-3. Let us assume the sender has decided
Salt & Pepper Noise
A i, j
Secret Data Binary
Sequence Encry ption
Encry pted Secret Data
to engraft all three available set of pixe ls in a binding image with the follo wing order: salty pixe ls, peppery pixe ls, pure
Fig.2: Stego System Encoder Functional Module
Procedure done in stego system encoder has e xplicated in this division. As cited earlie r stego system encoder utilizes the proposed schemes, Noise guided stegging and Adaptive K- bit engrafting. In a preprocessing step Salt & Pepper Noise is added with the binding image. Then the three sets of pixe ls in the noisy image are used to lead the adaptive engrafting scheme for concealing the clandestine data into the binding image as shown in fig. 2.
Algorithm for Engrafting
Binding image (Aij)
Clandestine data bit stream (F)
Adaptive key for K-bits
1. Stego image (Oij)
Step-1: Encrypt the clandestine data (F) by offering a Symmetric key.
Step-2: Find the binary bit strea m of c landestine data (F) Step-3: Interpret the binding image (Aij) for concealment
Step-4: Add the Salt & Pepper noise with defined Density (Ex: 0.04, 0.06 etc) to the copy of binding image. Na me th is as Guid ing image (Bij)
Step-5: Acquire the Gu iding image and determine the pixe l sets,
If Bij ==0 then Sa lty pixe ls (Bs,ij),
Else, if Bij == 255, then Peppery pixels (Bp,ij) Else, if Bij == Aij then Pure pixels (Bu,ij)
Step-6: Done the engrafting through the decision making as
If the key for Bs,ij Bu,ij Bp,ij 0, then choose the entire binding image and separate them into three sets based on the pixe l va lues.
Else, if the key for Bs,ij Bu,ij 0 & Bp,ij = 0, then choose and separate the pixe ls of Sa lty & pure and leave the Peppery pixe ls in binding image.
Else, if the key for Bs,ij Bp,ij 0 & Bu,ij = 0, then choose and separate the pixe ls of Sa lty & Peppery and leave
Step-10: Represent the resultant data engrafted image as Stego image (Oij)
Algorithm for clandestine data retrieving
Stego image (Oij)
Gu iding image (Bij)
Adaptive K-bit key
1. Retrieved clandestine data (F)
Step-1: Interpret the stego image (Oij) Step-2: Interpret the Gu ided image (Bij)
Step-3: Find and separate the emplace ment of pixe l sets such as Salty, Peppery and pure.
Step-4: Enter the same keys for k1, k2, k3 as entered in the engrafting step
Step-5: Clandestine data retrieving:
If k1 is 3 then retrieve three M SBs of Clandestine data (F) fro m the LSBs of Stego image (Oij)
Else, if k1 is 2 then retrieve 2bits of F
Else, if k1 is 1 then retrieve 1bit o f F
Else, if k1 is 0 then no data has engrafted into that corresponding pixel.
Step-6: Repeat the step-5 for k2 & k3.
Step-7: Co mbine the data bits retrieved fro m k1, k2 and k3. Ste9-8: Convert the retrieved bits into characters.
Step-9: Dec rypt the retrieved data by providing the same key as used in encryption.
TEST ING MEASURES
Bits Per Pixels (BPP)
The principal target of this paper is to attain eminent concealing capacity over the single binding image. Th is engrafting capacity is amended by number of bits engrafted into single pixe l. This is assessed as follows,
BPP = () (1)
the Pure pixe ls in b inding image.
Else, if the key for Bp,ij Bu,ij 0 & Bp,ij = 0, then choose and separate the pixe ls of peppery & pure and leave the Salty pixe ls in b inding image.
Step-7: Let us assume all the three pixel sets are chosen for Adaptive K-bit engrafting. Then the ma ximu m key
= total nu mber of b its engrafted
= M * N
M= Nu mber o f pixe ls in ro w of 2D image N= Nu mber of pixels in colu mn of 2D image
possibilit ies are as follows,
k1 = key for peppery pixe ls k2 = key for pure pixe ls
k3 = key for salty pixels
K= k1, k2, k3; K alte rs fro m 1 to 3.
Mean Square Error (MSE)
It is the measure of divergence between the input binding image pixe ls (Aij) and consequent stego image pixe ls (Oij). A a mend system must have lowest MSE.
K= 1,2,3; 1,3,2; 2,1,3; 2,3,1; 3,1,2; 3,2,1
MSE = 1
Step-8: Choose another binding image if the size is not enough to engraft the entire clandestine data bit streams.
Step-9: Engraft the MSBs of clandestine data bit streams into
the LSBs of binding image as mentioned in steps 6&7.
M= Nu mber o f pixe ls in ro w of 2D image
N= Nu mber of pixels in colu mn of 2D image
Peak Signal to Noise Ratio (PSNR)
It is the measure of exa mining the lineament of the stego image. A a mend system must have more e minent PSNR. The system with PSNR around 45-50d B is believed as good system. A system with PSNR above 50dB is conceived as much quality system for a steganography technique.
Binding image size: 35.6kb , Clandestine data size:
Adaptive K-bit Clandestine
Total No. of bits Engrafted
Tables 2 to 5, de monstrates that the proposed scheme has the e xtre me ly high data concealing capacity. Adaptive algorithm assures that the quality of the resultant stego image is not compensated.
TABLE.2: PERFORMANCE MEASURES FOR BINDING IMAGE TEMPLE
PSNR = 1010
( ) dB (3)
= Ma ximu m intensity of 2D image.
RESULT S & DISCU SSION
The poposed Noise Guided Random Stegging with
adaptive K- bit engrafting Stego system has been enforced in four unlike binding images. The Concealing Capacity of the stego images has been assessed and the consequences are evidenced in tables 2-5. Init ially, the stego image concealing capacity was gauged by the simple LSB substitution with standardized key engrafting in all the pixe ls and the consequences are e xhib ited in table .1. To establish the enhanced concealing capacity and quality of stego image developed by the proposed approach, the estimated BPP, Total engrafting capacity, MSE and PSNR of the stego image are co mpared with the results presented in table.1.
In table.1 it can be noticed that the concealing capacity and BPP are raised fro m k=1 to k=4, but values of MSE and PSNR dimin ished respectively. He re the c landestine data with the size of 71.5kb is engrafted into Temple and baboon binding images.
This fluctuation should not be the case for a amend stego system. A good system must have high concealing capacity as well as superiority stego image. This retreat in the existing simp le LSB substitution with standardized key engrafting can be defeat by employing the proposed Adaptive K- bit Engraft ing technique.
Number of Clandestine Data bits
of Bits Engrafted
Total No. of Bits
TABLE.1: BPP, MSE, PSNR, CONCEALING CAPACITY OF EXISTING
TABLE.3: PERFORMANCE MEASURES FOR BINDING IMAGE BABOON
Binding image size: 192kb , Clandestine data size: 71.5kb
Total No. of
TABLE.4: PERFORMANCE MEASURES FOR BINDING IMAGE LENA
Binding image size: 15.6kb , Clandestine data size: 71.5kb
Clandestine data K= k1-k2-k3
of bits Engrafted
TABLE.5: PERFORMANCE MEASURES FOR BINDING IMAGE GANDHI
Binding image size: 5.96kb , Clandestine data size: 71.5kb
Clandestine data K= k1-k2-k3
of bits Engrafted
The images of the representing binding images with guiding and stego images are shown in figures 3 to 6. It can be recognized that there is no visual difference between the resultant image and the binding image.
Binding Guided Stego
Fig.3: Results by applying T emple image
Binding Guided Stego
Fig.4: Results by applying Baboon image
Binding Guided Stego
Fig.5: Results by applying Lena image
Binding Guided Stego
Fig.6: Results by applying Gandhi image
In all the above images Sa lt & Pepper noise is added with the density of 0.04. This density is user defined based on the require ment. If a sender demands more pixels with modified intensity then the density can be raised.
SECURIT Y ANALYSIS
The proposed healthy stego system has mu ltilayer shelter against different attacks. For each module, the proposed technique that leads in the ma ximu m BPP, minimu m MSE and good PSNR values is adopted here, there by augmenting the concealing capacity of the s tego image. Moreover, splicing the binding image with three dissimila r pixe l sets by using noise guided image me liorates the
protection of the vital message because only the authorized user has the key to the correct compounding of data set pixe l emp lace ments and binary pattern applied in each combination. Ultimately, by imparting fle xib ility in the choice of terminus a quo of the engrafting pixe l set, this method assures huge security and invulnerability.
The paces necessitated in the protection mechanism a re as follows:
Step-1: The confidentia l informat ion is encrypted using symmetric key.
Step-2: Noise guided stegging technique is employed
Step-3: Three unlike pixe ls sets of the binding image are obtained for engrafting.
Step-4: Adaptive K- bit engrafting technique builds different
binary pattern for concealment.
Step-1 furnishes a cryptic effect for the confidential data before engrafting, which contributes to the first layer of security. Step-2 furnishes a retiring platform for engrafting the informat ion. Step-3 contributes to choosing the pixe ls sets for engrafting, clandestine data need not to be engrafted in all the pixels sets, thus increasing the protection. Step -4 e xtends to furnish unlike binary pattern for apiece pixe l sets, it guarantees that hacking the data is unimaginable .
The noise guided stegging with adaptive K- bit engrafting techniques enforced in this paper e mploying an intelligent helter-skelter engraft ing process for the encrypted clandestine data. From the computed results of the concealing capacity of the stego image generated by the proposed method, it is noticed that the adaptive K- bit engrafting techniques supplants the existing techniques in meliorating the concealing capacity of the stego image. Furthermo re t he noise guided stegging technique extends significantly improved security without markedly conciliatory the payload. In addition, choice of engrafting adaptive key and the pixe ls sets are allowed for user defined decision making instead of manual and predefined decisions. This facilitates the system more user friendly. Moreover the noise employed in the proposed technique is random; it leads the classificat ion of pixe l sets and its positions in the binding image are unpredictable.
Authors wish to thank Mrs. Devipriya sila mbarasan/ PG Scholar/ SASTRA University for her time , linguistic & technical support.
Digital image steganography: Survey and analysis of current methods Abbas Cheddad, JoanCondell, KevinCurran,
PaulM cKevitt, journal of signal processing
Exploring steganography: seeing the unseen, N.F.Johnson, S.Jajodia, IEEE Computer 31 (2) (1998) 2634.
Steganography: past, present, future. J.C.Judge, SANS Institute publication, <http: // www.sans.org/reading_room/ whitepapers/ stenganography/ 552.php>, 2001
Hide and seek: an introduction to steganography, N.Provos, P.Honeyman, IEEE Security and Privacy 1(3) (2003) 3244.
Data-hiding codes, P.M oulin, R.Koetter, Proceedings of the IEEE 93 (12) (2005) 20832126.
Cryptography: current status and future trends, S.B.Sadkhan,
in: Proceedings of IEEE International Conference on Information& Communication Technologies: From Theory to Applications,
Damascus. Syria, April1923, 2004, pp.417418
Steganalysis using higher-order image statistics, S.Lyu, H.Farid, IEEE Transactions on Information Forensics and Security 1(1) (2006)111119.
The code breakers: the comprehensive history of secret communication from ancient times to the Internet, D.Kahn, Scribner, December 5, 1996.
Information noyee, information cach, J.P.Delahaye, Pour la
Science 229 (1996) 142146 /www.apprendre-en- ligne.net/crypto/stega no/229_142_146.pdfS (in French).
The prisoners problem and the subliminal channel , G.J.Simmons, in: Proceedings of International Conference on Advances in Cryptology, CRYPTO83, August2224, 1984, pp.51 67.
A cautionary note on image down grading, C.Kurak,
J.M cHugh, in: Proceedings of the IEEE 8th Annual Computer Security Applications Conference, 30 November4 December, 1992, pp 153-159.
Discovering hidden evidence, C.Hosmer, Journal of Digital Forensic Practice (1)(2006)4756.
Applications for data hiding , W.Bender, W.Butera, D.Gruhl, R.Hwang, F.J.Paiz, S.Pogreb, IBM Systems Journal 39 (3&4) (2000) 547568.
Introduction to information hiding, F.A.P.Petitcolas, in: S.Katzen beisser, F.A.P. Petitcolas (Eds.), Information Hiding
Techniques for Steganography and Digital Watermarking, Artech House, Inc., Norwood, 2000.
A secure data hiding technique with heterogeneous data- combining capability for electronic patient records, S.M iaou, C.Hsu, Y.Tsai, H.Chao, in: Proceedin gs of the IEEE 22nd Annual EM BS International Conference, Chicago, USA, July2328, 2000, pp. 280283.
Water marking medical images with patient information , U.C.Nirinjan, D.Anand, in: Proceedings of the 20th Annual International Conference of the IEEE Engineering in M edicine and Biology Society, Hong Kong,China,29October 1November1998,pp.703706.
Protection of mammograms using blind steganography and water marking, Y.Li, C.Li, C.Wei, in: Proceedings of the IEEE International Symposium on Information Assurance and Security, 2007, pp.496499.
Using extended file information (EXIF) file headers in digital evidence analysis, P.Alvarez, International Journal of Digital Evidence, Economic Crime Institute (ECI) 2 (3) (2004) 15
A Novel Steganographic Method for Gray-Level Images With four-pixel Differencing and LSB Substitution- M .B. Ould M EDENI, 978-1-61284-732-0/11/$26.00, 2010 IEEE.
A New Approach for LSB Based Image Steganography using Secret Key- S. M . M asud Karim, M d. Saifur Rahman, M d. Ismail Hossain in: Proceedings of 14th International Conference on Computer and Information Technology (ICCIT 2011)
A more secure steganography based on adaptive pixel-value differencing scheme-Weiqi Luo Â· Fangjun Huang Â· Jiwu Huang, Springer Science+Business M edia, LLC 2010.
A Novel Keyless Algorithm for Steganography- Supriya Rai and Ruchi Dubey, in: IEEE proceedings 2012.
Who decides hiding capacity? I, the Pixel Intensity – Rengarajan Amirtharajan, K.Ramkrishnan, M .Vivek Krishna, Nandhini.J and John Bosco Balaguru Rayappan, in: IEEE proceedings 2012.
An intelligent chaotic embedding approach to enhance stego- image Quality- Rengarajan Amirtharajan, John Bosco Balaguru Rayappan- Journal of information science.
Space-Filling Curves, Hans Sagan, Springer- Verlag, New
York, 1994. ISBN: 0-387-94265-3.
A. Westfeld Space filling curves in steganalysis, in: E.J. Delp, III, P.W. Wong (Eds), Security, Steganography and Watermarking of M ultimedia Contents VII SPIE 5681, 2005, pp. 2837.
Information secured by the guidance of salt & pepper noise-
Jayaseelan. J, Kruthika. B- 2013 IEEE International Conference on Computational Intelligence & Computing Research.