Research Article A Formula Adaptive Pixel Pair Matching Steganography Algorithm Min Long 1,2 and Fenfang Li 1 1 College of Computer and Communication Engineering, Changsha University of Science and Technology, 410114, China 2 Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on Transportation, Changsha University of Science and Technology, Changsha, Hunan Province 410114, China Correspondence should be addressed to Min Long; [email protected] Received 23 January 2018; Revised 31 March 2018; Accepted 30 April 2018; Published 5 July 2018 Academic Editor: Mehdi Hussain Copyright © 2018 Min Long and Fenfang Li. is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Pixel pair matching (PPM) is widely used in digital image steganography. As an important derivation, adaptive pixel pair matching method (APPM) offers low distortion and allows embedded digits in any notational system. However, APPM needs additional space to store, calculate, and query neighborhood set, which needs extra cost. To solve these problems, a formula adaptive pixel pair matching (FAPPM) method is proposed in this paper. e basic idea of FAPPM is to use the formula to get the stego image pixel pair without searching the neighborhood set for the given image pixel pair. is will allow users to embed secret message directly without storing and searching the look-up table. Experimental results and analysis show that the proposed method could embed secret data directly without searching the neighborhood sets by using a formula and it still maintains flexibility in the selection of notional system, high image quality, and strong anti-steganalysis ability. 1. Introduction Information hiding is a technology of embedding secret data into the media for covert communication [1]. With the rapid development of Internet, a large number of data are transmitted over the Internet. At present, the main media using for data hiding includes images, audio, and video, where digital image is the most widely used media [2]. Researchers have shown a great interest in image steganography for the last decade [3]. LSB replacement [4] is one of the most commonly used steganographic techniques, which makes full use of the characteristics that the human visual system is not sensitive to small changes in pixels and the negligible contribution of the low bit plane of the pixel to the image quality. However, this method can only add 1 or remain unchanged for the even pixels and can only decrease 1 or remain unchanged for the odd pixels. erefore, this unbalanced embedding distortion leads to the histogram attack to the images [5, 6]. Chan et al. [7] proposed an optima pixel adjustment process (OPAP) method, which adjusted the pixels to reduce the distortion caused by least significant bit (LSB) embedding. e LSB and OPAP methods both employed one pixel as an embedding unit to embed secret message. As the development of steganography, methods using two or more pixels as a basic unit for B-ary secret information embedding were put forward. is kind of stenographic algorithm can improve the embedding capacity and image quality by subtle modifying the pixel. In 2006, Miekikainen [11] proposed a LSB matching method. It employed two pixels as embedding unit. In this method, when payload was 1 bit per pixel, the mean square error (MSE) is 0.375, while the MSE of LSB [4] was 0.5. Zhang and Wang [12] proposed exploiting modification direction (EMD) method, which added and subtracted 1 in one pixel and embedded 2n + 1-ary secret message in n pixels. When n = 2, a quinary number was embedded in each pair of pixels. e capacity can reach the maximum (1/2)log 2 5 = 1.161 bit per pixel (bpp). Chao et al. [13] extended this method and proposed a diamond encoding (DE) method. It can embed 2k 2 +2k+1-ary information to each pair of pixels and achieve high embedding efficiency by adding and subtracting 1 operation in n pixels. In [8], the author used a codebook to Hindawi Advances in Multimedia Volume 2018, Article ID 7682098, 8 pages https://doi.org/10.1155/2018/7682098
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Research ArticleA Formula Adaptive Pixel Pair MatchingSteganography Algorithm
Min Long 12 and Fenfang Li1
1College of Computer and Communication Engineering Changsha University of Science and Technology 410114 China2Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on Transportation Changsha University of Science andTechnology Changsha Hunan Province 410114 China
Correspondence should be addressed to Min Long caslongmaliyuncom
Received 23 January 2018 Revised 31 March 2018 Accepted 30 April 2018 Published 5 July 2018
Academic Editor Mehdi Hussain
Copyright copy 2018 Min Long and Fenfang Li This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited
Pixel pair matching (PPM) is widely used in digital image steganography As an important derivation adaptive pixel pair matchingmethod (APPM) offers low distortion and allows embedded digits in any notational system However APPM needs additionalspace to store calculate and query neighborhood set which needs extra cost To solve these problems a formula adaptive pixelpairmatching (FAPPM)method is proposed in this paperThe basic idea of FAPPM is to use the formula to get the stego image pixelpair without searching the neighborhood set for the given image pixel pair This will allow users to embed secret message directlywithout storing and searching the look-up table Experimental results and analysis show that the proposed method could embedsecret data directly without searching the neighborhood sets by using a formula and it still maintains flexibility in the selection ofnotional system high image quality and strong anti-steganalysis ability
1 Introduction
Information hiding is a technology of embedding secretdata into the media for covert communication [1] With therapid development of Internet a large number of data aretransmitted over the Internet At present the main mediausing for data hiding includes images audio and videowheredigital image is the most widely used media [2] Researchershave shown a great interest in image steganography for thelast decade [3] LSB replacement [4] is one of the mostcommonly used steganographic techniques whichmakes fulluse of the characteristics that the human visual system isnot sensitive to small changes in pixels and the negligiblecontribution of the low bit plane of the pixel to the imagequality However this method can only add 1 or remainunchanged for the even pixels and can only decrease 1or remain unchanged for the odd pixels Therefore thisunbalanced embedding distortion leads to the histogramattack to the images [5 6] Chan et al [7] proposed an optimapixel adjustment process (OPAP) method which adjustedthe pixels to reduce the distortion caused by least significant
bit (LSB) embedding The LSB and OPAP methods bothemployed one pixel as an embedding unit to embed secretmessage As the development of steganography methodsusing two or more pixels as a basic unit for B-ary secretinformation embedding were put forward This kind ofstenographic algorithm can improve the embedding capacityand image quality by subtle modifying the pixel
In 2006 Miekikainen [11] proposed a LSB matchingmethod It employed two pixels as embedding unit In thismethod when payload was 1 bit per pixel the mean squareerror (MSE) is 0375 while theMSE of LSB [4] was 05 Zhangand Wang [12] proposed exploiting modification direction(EMD) method which added and subtracted 1 in one pixeland embedded 2n + 1-ary secret message in n pixels When n= 2 a quinary number was embedded in each pair of pixelsThe capacity can reach the maximum (12)log25 = 1161bit per pixel (bpp) Chao et al [13] extended this methodand proposed a diamond encoding (DE) method It canembed 2k2+2k+1-ary information to each pair of pixels andachieve high embedding efficiency by adding and subtracting1 operation in n pixels In [8] the author used a codebook to
HindawiAdvances in MultimediaVolume 2018 Article ID 7682098 8 pageshttpsdoiorg10115520187682098
2 Advances in Multimedia
Table 1 Extraction Function Coefficient 119888119861 of APPM
1198882 1198883 1198884 1198885 1198886 1198887 1198888 1198889 11988810 11988811 11988812 11988813 11988814 11988815 119888161 1 2 2 2 2 3 3 3 3 4 5 4 4 611988817 11988818 11988819 11988820 11988821 11988822 11988823 11988824 11988825 11988826 11988827 11988828 11988829 11988830 119888314 4 4 8 4 5 5 5 5 10 5 5 5 12 1211988832 11988833 11988834 11988835 11988836 11988837 11988838 11988839 11988840 11988841 11988842 11988843 11988844 11988845 119888467 6 6 10 15 6 16 7 7 6 12 12 8 7 711988847 11988848 11988849 11988850 11988851 11988852 11988853 11988854 11988855 11988857 11988858 11988859 11988860 11988861 119888627 7 14 14 9 22 8 12 21 16 24 22 9 8 811988863 1198886414 14
(a) Φ4 1198884 = 2 (b) Φ5 1198885 = 2 (c) Φ6 1198886 = 2 (d) Φ9 1198889 = 3 (e) Φ13 11988813 = 5
(f) Φ16 11988816 = 6 (g) Φ25 11988825 = 5
Figure 1 Neighborhood set (shaded region) for APPM
improve the EMD scheme and one secret (2n+x-1)-ary digitwas hidden in a group of pixels in an image as a modifiedsecret digit In [9] the authors proposed a method to modifya group of pixels by plusmn1 to embed a secret digit but it isonly applicable to 3n-ary notational system Kuo et al [14]proposed a formula diamond encoding (FDEMD) data hidescheme and it could conceal a digit in (2k2+2k+1)-ary systemIt simplified the embedding procedure and embedded secretdata without storing and calculating characteristic valuematrix Hong et al [10] designed a new extraction functionand new neighborhood set of two pixels called adaptivepixel pair matching (APPM) It allowed embedding digitsin arbitrary notational system and the distortion caused
by embedment using APPM was minimized therefore theresultant marked image quality could be well preserved[15] In [16] secure adaptive pixel pair matching (SAPPM)was proposed to hide multiple data types such as textimage and audio which incorporated cryptography alongwith steganography A transformed version of adaptive pixelpair matching (APPM) was used for image steganography toget lower distortion [17] However APPM need to calculatestore and query the modified neighborhood set table
Based on the above methods this paper simplifies theembedding procedure and designs an extraction function toconstruct a formula adaptive pixel pair matching (FAPPM)method It does not need to calculate store and query the
Advances in Multimedia 3
Y
NCompute the embedding process
Are there remaining
pixelsEndGet pixelStart secret data s pair (x1x2)
Input cB and
Figure 2 The embedding process
Input A pixel pair (1199091 1199092) extraction function coefficient 119888119861 and secret data 119904Output Stego pixel pair (11990910158401 11990910158402)Step 1 Set 119891 = (1199091 + 1198881198611199092)mod119861Step 2 Set 119896 = lceil(lceilradic119861rceil minus 1)2rceilStep 3 Set119863 = 119904 minus 119891Step 4 If119863 lt 0 then119863 = 119863 + 119861Step 5 Set 119899119890119909119905 1199051 = |119863|mod 119888119861Step 6While 119894 = 1 to 4 do
Set 1199051 = 119899119890119909119905 1199051Set 1199052 = (119863 minus 1199051)119888119861If |1199051| le 119896ampamp|1199052| le 119896 then
Set 11990910158401 = 11990910158401 + 1199051Set 11990910158402 = 11990910158402 + 1199052
ElseSwitch (119894)
Case 1Set 119899119890119909119905 1199051 = 1199051 minus 119888119861
Case 2Set119863 = 119863 minus 119861Set 119899119890119909119905 1199051 = minus(|119863|mod 119888119861)
Case3Set 119899119890119909119905 1199051 = 1199051 + 119888119861
Case4Print ldquoErrorrdquo
End SwitchEnd if
End While
Algorithm 1
modified neighborhood set table and it can realize the datahiding in any notional system
2 A Review of Adaptive Pixel PairMatching (APPM)
The APPM method [10] used a pair of pixels (119909 119910) as acoordinate where an extraction function 119891119860119875119875119872(119909 119910) wasdesigned Then a neighborhood set Φ(119909 119910) of (119909 119910) wasestablished
119891119860119875119875119872 (119909 119910) = (119909 + 119888119861119910) mod119861 (1)
where 119891(119909 119910) and Φ(119909 119910) satisfied the following threeconditions
(i) In the neighborhood set Φ(119909 119910) there are exactly 119861pairs of coordinates
(ii) In the neighborhood set Φ(119909 119910) the extracted func-tion values for each coordinate aremutually exclusive
(iii) According to 119891(119909 119910) and Φ(119909 119910) a digit can beembedded in any notional system
The way to find the extraction function coefficient 119888119861and Φ(119909 119910) can be converted to find the following optimalsolution
Minimizesum119861minus1119894=0 [(119909119894minus119909)2+(119910119894minus119910)2] subject to119891(119909119894 119910119894) isin0 1 119861 minus 1 where 119891(119909119894 119910119894) = 119891(119909119895 119910119895) if 119894 = 119895 and 0 le119894 119895 le 119861 minus 1According to the above 119888119861 and Φ(119909 119910) can be calculated
with different B-ary For APPM proposed by Hong [10] 119888119861corresponding to B-ary is listed in Table 1 Meanwhile partsof Φ(119909 119910) corresponding to B-ary are illustrated in Figure 1
Compared with DE and EMD method APPM has theflexibility to choose a better notational system for dataembedding to decrease the image distortion The selection
4 Advances in Multimedia
Input stego image 119878Output Secret dataStep 1 Divide the stego image 119878 into non overlapping pixel pairs (1199091015840119894 1199101015840119894 )Step 2 Calculate 119904119894 = 119891(1199091015840119894 1199101015840119894 ) = (1199091015840119894 + 1198881198611199101015840119894 )mod119861 where 119894 represents the i-th pixel pairStep 3 Calculate all 119904119894 and convert them to binary stream119898
Algorithm 2
(a) Lena (b) Barbara (c) Pepper (d) Boat
(e) Tiffany (f) Baboon (g) Zelda (h) Airplane
Figure 3 The eight gray cover images
(a) Lena (b) Barbara (c) Pepper (d) Boat
(e) Tiffany (f) Baboon (g) Zelda (h) Airplane
Figure 4 The eight stego images (B=27 PSNR=45dB)
Advances in Multimedia 5
B=5B=9B=13B=17
B=21B=25B=41
46
48
50
52
54
56
58
60
62
64
PSN
R (d
B)
02 03 04 05 06 07 08 09 101Embedding Capacity ()
Figure 5 The relationships between embedding payload and image quality
of B-ary system is determined by the size of the coverimage C Given the size of C is MtimesN B is the minimumvalue satisfying lfloor119872 times 1198732rfloor ge |119904119861| However it needed tocalculate store and query the neighborhood set as shown inFigure 1
3 The Proposed Formula Adaptive Pixel PairMatching Method (FAPPM)
In order to solve the above shortcomings this paper putsforward a formula adaptive pixel pair matching embeddingmethod to find the stego-pixel pair without a neighborhoodset
31 Embedding Procedure In the embedding procedure fourvectors at most are produced Two vectors are calculatedwhen Dgt0 and the other two vectors are calculated whenDlt0 In Algorithm 1 i represents vectors 1 to 4 in turnFigure 2 shows the embedding process overview
Example 1 For a cover pixels pair (5 6) secret data 119904 = 8(16)and extraction function coefficient 11988816 = 6 the stego imagepixels pair (11990910158401 11990910158402) = (4 6) is obtained by using Algorithm 1
Step 1 Calculate 119891 = (5 + 6 times 6)mod 16 = 9 119896 = lceil(lceilradic119861rceil minus1)2rceil = 2Step 2 Calculate119863 = 119904 minus 119891 = minus1 As119863 lt 0119863 = 119863 + 119861 = 15is obtained
Step 3 Calculate 119899119890119909119905 1199051 = |119863|mod 11988816 = 3(1) Round 1 1199051 = 3 1199052 = 2(2) |1199051| gt 119896ampamp|1199052| gt 119896 then 119899119890119909119905 1199051 = 1199051 minus 119888119861 = minus3(3) Round 2 1199051 = minus3 1199052 = 3(4) |1199051| gt 119896ampamp|1199052| gt 119896 then 119863 = 119863 minus 119861 = minus1 119899119890119909119905 1199051 =minus(|119863|mod 11988816) = minus1
(5) Round 3 1199051 = minus1 1199052 = 0(6) |1199051| le 119896ampamp|1199052| le 119896 then return (4 6)
32 Extraction Procedure Through extraction functionsecret digits can be extracted from the stego image Thedetailed process is given in Algorithm 2
33 Overflow Problem and Solution If an overflow or under-flow problem occurs that is (1199091015840 1199101015840) lt 0 or (1199091015840 1199101015840) gt 255a nearest (11990910158401015840 11991010158401015840) should be found in the neighborhood of(119909 119910) such that 119891(11990910158401015840 11991010158401015840) = 119904119861 This can be done by solvingthe optimization problem
Minimize (119909 minus 11990910158401015840)2 + (119910 minus 11991010158401015840)2 Subject to 119891 (11990910158401015840 11991010158401015840) = 119904119861 0 le 11990910158401015840 11991010158401015840 le 255 (2)
4 Experimental Results and Analysis
41 Experimental Results The experiments are performedusing Matlab R2013a and eight 512 times 512 grayscale imagesare used as shown in Figure 3 The stego images are shown inFigure 4 where B=27
As seen from Figures 3 and 4 the difference between thecover images and the corresponding stego images is very littleand cannot be distinguished byhumanrsquos eyes It illustrated thegood imperceptibility of the proposed method
As message embedding it will introduce the distortion inthe image Peak signal-to-noise ratio (PSNR) is usually usedto measure the quality of image The definition of PSNR is asfollows
119875119878119873119877 = 10 times log10 ( 2552119872119878119864) (3)
6 Advances in Multimedia
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
HorizontalVertical
(a) FAPPM B=53 D=134
HorizontalVertical
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
(b) FDEMD B=53D=936
HorizontalVertical
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Occ
urre
nce
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
(c) FAPPM B=221 D=189
HorizontalVertical
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
0
500
1000
1500
2000
2500
3000
3500
4000
4500O
ccur
renc
e
(d) FDEMD B=221 D=1449
Figure 6 Comparison of the averaged vertical and horizontal difference histograms of FAPPM and FDEMD
whereMSE is themean square error between the cover imageand stego image it is defined as follows
119872119878119864 = 1119872 times 119873119872sum119894=0
119873sum119895=0
(119901119894119895 minus 1199011015840119894119895)2 (4)
Here the symbols 119901119894119895 and 1199011015840119894119895 represent the pixel valuesof the cover image and stego image in the position (119894 119895)respectively and 119872 and 119873 are the width and height of theoriginal image
As the proposed method can embed secret digit in anynotional system experiments are done to test the relationshipbetween embedding payload and image quality and the
results are shown in Figure 5 It can be found that the PSNR isdecreased as the embedding capacity is increased Howeverthe PSNR still achieved a high value when the embeddingcapacity reached 1
42 Comparison with OtherMethods Here EMD [8] EMD-3[9] APPM and FAPPM are compared from six aspects theembedding method the national system payload capacityPSNR and the storage space The results are listed in Table 2As seen from Table 2 FAPPM method uses a mathematicalmethod to embed secret data and it does not need any spaceto store neighbor table furthermore it does not affect thecapacity and image quality
Advances in Multimedia 7
Table 2 Comparison of results
Contents of comparison EMD[8] EMD-3[9] APPM[10] Proposed FAPPMEmbedding method Matrix and search Matrix and search table look-up Mathematic methodNotational systems of B-ary fixed fixed arbitrary arbitraryPayload (bpp) B=25 2471 2471 232 232PSNR (dB) 439 429 481 481Need the storage space Yes Yes Yes No
Figure 7 The difference of Rm and R-m for RS attack
Figure 8 The difference of Sm and S-m for RS attack
43 Analysis of the Security Anti-steganalysis is one of themost important criteria to measure the performance of asteganographic method In this paper a detection methodbased on histogramdifferential statistics analysis proposed byZhao [18] is used to test the security of the FAPPM methodNormally in an image with no hidingmessage the horizontaldifference histogram ℎ and the vertical difference histogramV are coincident But when the message is embedded in apair of pixels its ℎ and V will be changed The distancebetween ℎ and V is used to construct a statistical detector
to detect the variation between histograms The distance isdefined as follows
119863 = ( 2119879sum119894=minus2119879
(ℎ (119894) minus V (119894)))12
(5)
where 119879 is a predefined threshold and 119863 represents thedifference between ℎ and VThe larger the119863 is the greaterthe difference between ℎ and V is That is the probabilitythat the image contains secret information is high Hereexperiments are done to compare the histogram variationof FAPPM and FDEMD under high payload Both FAPPMand FDEMDmethods are used to generate 100 stego imagesrespectively ℎ V and their average value are calculatedrespectively The parameters are B=53 B=211 and T=20 Allthe test images were fully embedded The experiment resultsare shown in Figure 6 It can be seen that there is almostno difference between ℎ and V for FAPPM while that forFDEMD is significant which indicates the probability thatthe successful steganalysis for FDEMD is higher than that ofthe proposed method
The RS attack method can detect LSB secret data embed-ding in grayscale or color images Each pixel block is classifiedinto the regular group 119877 the singular group 119878 and theunusable group 119880 by a flipping function and mask 119872 119877119898119878119898 and 119880119898 denote the number of 119877 119878 and 119880 respectivelyFor inverse mask -119872 119877-119898 119878-119898 and U-119898 denote thenumber of 119877 119878 and 119880 respectively When no information isembedded119877119898 -119877-masymp0 and 119878119898 -119878-masymp0TheRS attack resultsare shown in Figures 7 and 8 It can be seen that the algorithmof this paper can guarantee 119877119898 -119877-masymp0 and 119878119898 -119878-masymp0 andthe existence of secret information cannot be detected by RSsteganalysis method
5 Conclusion
This paper proposed a simple and convenient data embed-ding method based on APPM Compared with the APPMmethod it has the advantage of no needing to compute andstore the neighborhood set Compared with the FDEMDmethod the secret data of any notional system is realized bythe FAPPMmethod which makes the embedding notationalsystem selection more flexible The experimental resultsshowed that FAPPM method has high image quality andthe strong anti-steganalysis ability Our future work willbe concentrated on the use of the formula method of theadjacent three pixels as the embedding unit
8 Advances in Multimedia
Data Availability
The data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
The authors declare that they have no conflicts of interest
Acknowledgments
This work was supported in part by project supported byNational Natural Science Foundation of China (Grant no61572182 no 61370225) and project supported by HunanProvincial Natural Science Foundation of China (Grant no15JJ2007)
References
[1] J Fridrich Steganography in Digital Media Principles Algo-rithms and Applications Cambridge University Press NewYork NY USA 2009
[2] A Cheddad J Condell K Curran and P Mc Kevitt ldquoDigitalimage steganography survey and analysis of current methodsrdquoSignal Processing vol 90 no 3 pp 727ndash752 2010
[3] M Hussain A W Wahab Y I Idris A T Ho and K JungldquoImage steganography in spatial domain A surveyrdquo SignalProcessing Image Communication vol 65 pp 46ndash66 2018
[4] N Provos and P Honeyman ldquoHide and seek an introduction tosteganographyrdquo IEEE Security amp Privacy vol 1 no 3 pp 32ndash442003
[5] J Fridrich M Goljan and R Du ldquoReliable Detection of LSBSteganography in Color and Grayscale Imagesrdquo The Workshopon Multimedia amp Security New Challenges ACM pp 22ndash282002
[6] A D Ker ldquoSteganalysis of LSB matching in grayscale imagesrdquoIEEE Signal Processing Letters vol 12 no 6 pp 441ndash444 2005
[7] C-K Chan and L M Cheng ldquoHiding data in images by simpleLSB substitutionrdquo Pattern Recognition vol 37 no 3 pp 469ndash474 2004
[8] C Kim ldquoData hiding by an improved exploiting modificationdirectionrdquoMultimedia Tools and Applications vol 69 no 3 pp569ndash584 2014
[9] X Niu M Ma R Tang and Z Yin ldquoImage steganography viafully exploiting modification directionrdquo International Journal ofSecurity and Its Applications vol 9 no 5 pp 243ndash254 2015
[10] W Hong and T-S Chen ldquoA novel data embedding methodusing adaptive pixel pair matchingrdquo IEEE Transactions onInformation Forensics and Security vol 7 no 1 pp 176ndash1842012
[11] J Mielikainen ldquoLSB matching revisitedrdquo IEEE Signal ProcessingLetters vol 13 no 5 pp 285ndash287 2006
[12] X Zhang and S Wang ldquoEfficient steganographic embeddingby exploiting modification directionrdquo IEEE CommunicationsLetters vol 10 no 11 pp 781ndash783 2006
[13] R Chao H Wu C Lee and Y Chu ldquoA Novel Image DataHiding Scheme with Diamond Encodingrdquo EURASIP Journal onInformation Security vol 2009 no 1 p 658047 2009
[14] W-C Kuo P-Y Lai C-C Wang and L-C Wuu ldquoA formuladiamond encoding data hiding schemerdquo Journal of Information
Hiding and Multimedia Signal Processing vol 6 no 6 pp 1167ndash1176 2015
[15] W Hong M Chen T Chen and C Huang ldquoAn efficientauthentication method for AMBTC compressed images usingadaptive pixel pair matchingrdquoMultimedia Tools amp Applicationsvol 77 no 4 pp 4677ndash4695 2018
[16] T Edwina Alias D Mathew and A Thomas ldquoSteganographicTechnique Using Secure Adaptive Pixel Pair Matching forEmbedding Multiple Data Types in Imagesrdquo in Proceedings ofthe 5th International Conference on Advances in Computing andCommunications ICACC 2015 pp 426ndash429 India September2015
[17] J Pappachan and J Baby ldquoTransformed adaptive pixel pairmatching technique for colour imagesrdquo in Proceedings of theInternational Conference on Control Instrumentation Commu-nication and Computational Technologies ICCICCT 2015 pp192ndash196 India December 2015
[18] H Zhao H Wang and M Khurram Khan ldquoStatistical analysisof several reversible data hiding algorithmsrdquo Multimedia Toolsand Applications vol 52 no 2-3 pp 277ndash290 2011
International Journal of
AerospaceEngineeringHindawiwwwhindawicom Volume 2018
RoboticsJournal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Active and Passive Electronic Components
VLSI Design
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Shock and Vibration
Hindawiwwwhindawicom Volume 2018
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawiwwwhindawicom
Volume 2018
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018
Control Scienceand Engineering
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Journal ofEngineeringVolume 2018
SensorsJournal of
Hindawiwwwhindawicom Volume 2018
International Journal of
RotatingMachinery
Hindawiwwwhindawicom Volume 2018
Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
2 Advances in Multimedia
Table 1 Extraction Function Coefficient 119888119861 of APPM
1198882 1198883 1198884 1198885 1198886 1198887 1198888 1198889 11988810 11988811 11988812 11988813 11988814 11988815 119888161 1 2 2 2 2 3 3 3 3 4 5 4 4 611988817 11988818 11988819 11988820 11988821 11988822 11988823 11988824 11988825 11988826 11988827 11988828 11988829 11988830 119888314 4 4 8 4 5 5 5 5 10 5 5 5 12 1211988832 11988833 11988834 11988835 11988836 11988837 11988838 11988839 11988840 11988841 11988842 11988843 11988844 11988845 119888467 6 6 10 15 6 16 7 7 6 12 12 8 7 711988847 11988848 11988849 11988850 11988851 11988852 11988853 11988854 11988855 11988857 11988858 11988859 11988860 11988861 119888627 7 14 14 9 22 8 12 21 16 24 22 9 8 811988863 1198886414 14
(a) Φ4 1198884 = 2 (b) Φ5 1198885 = 2 (c) Φ6 1198886 = 2 (d) Φ9 1198889 = 3 (e) Φ13 11988813 = 5
(f) Φ16 11988816 = 6 (g) Φ25 11988825 = 5
Figure 1 Neighborhood set (shaded region) for APPM
improve the EMD scheme and one secret (2n+x-1)-ary digitwas hidden in a group of pixels in an image as a modifiedsecret digit In [9] the authors proposed a method to modifya group of pixels by plusmn1 to embed a secret digit but it isonly applicable to 3n-ary notational system Kuo et al [14]proposed a formula diamond encoding (FDEMD) data hidescheme and it could conceal a digit in (2k2+2k+1)-ary systemIt simplified the embedding procedure and embedded secretdata without storing and calculating characteristic valuematrix Hong et al [10] designed a new extraction functionand new neighborhood set of two pixels called adaptivepixel pair matching (APPM) It allowed embedding digitsin arbitrary notational system and the distortion caused
by embedment using APPM was minimized therefore theresultant marked image quality could be well preserved[15] In [16] secure adaptive pixel pair matching (SAPPM)was proposed to hide multiple data types such as textimage and audio which incorporated cryptography alongwith steganography A transformed version of adaptive pixelpair matching (APPM) was used for image steganography toget lower distortion [17] However APPM need to calculatestore and query the modified neighborhood set table
Based on the above methods this paper simplifies theembedding procedure and designs an extraction function toconstruct a formula adaptive pixel pair matching (FAPPM)method It does not need to calculate store and query the
Advances in Multimedia 3
Y
NCompute the embedding process
Are there remaining
pixelsEndGet pixelStart secret data s pair (x1x2)
Input cB and
Figure 2 The embedding process
Input A pixel pair (1199091 1199092) extraction function coefficient 119888119861 and secret data 119904Output Stego pixel pair (11990910158401 11990910158402)Step 1 Set 119891 = (1199091 + 1198881198611199092)mod119861Step 2 Set 119896 = lceil(lceilradic119861rceil minus 1)2rceilStep 3 Set119863 = 119904 minus 119891Step 4 If119863 lt 0 then119863 = 119863 + 119861Step 5 Set 119899119890119909119905 1199051 = |119863|mod 119888119861Step 6While 119894 = 1 to 4 do
Set 1199051 = 119899119890119909119905 1199051Set 1199052 = (119863 minus 1199051)119888119861If |1199051| le 119896ampamp|1199052| le 119896 then
Set 11990910158401 = 11990910158401 + 1199051Set 11990910158402 = 11990910158402 + 1199052
ElseSwitch (119894)
Case 1Set 119899119890119909119905 1199051 = 1199051 minus 119888119861
Case 2Set119863 = 119863 minus 119861Set 119899119890119909119905 1199051 = minus(|119863|mod 119888119861)
Case3Set 119899119890119909119905 1199051 = 1199051 + 119888119861
Case4Print ldquoErrorrdquo
End SwitchEnd if
End While
Algorithm 1
modified neighborhood set table and it can realize the datahiding in any notional system
2 A Review of Adaptive Pixel PairMatching (APPM)
The APPM method [10] used a pair of pixels (119909 119910) as acoordinate where an extraction function 119891119860119875119875119872(119909 119910) wasdesigned Then a neighborhood set Φ(119909 119910) of (119909 119910) wasestablished
119891119860119875119875119872 (119909 119910) = (119909 + 119888119861119910) mod119861 (1)
where 119891(119909 119910) and Φ(119909 119910) satisfied the following threeconditions
(i) In the neighborhood set Φ(119909 119910) there are exactly 119861pairs of coordinates
(ii) In the neighborhood set Φ(119909 119910) the extracted func-tion values for each coordinate aremutually exclusive
(iii) According to 119891(119909 119910) and Φ(119909 119910) a digit can beembedded in any notional system
The way to find the extraction function coefficient 119888119861and Φ(119909 119910) can be converted to find the following optimalsolution
Minimizesum119861minus1119894=0 [(119909119894minus119909)2+(119910119894minus119910)2] subject to119891(119909119894 119910119894) isin0 1 119861 minus 1 where 119891(119909119894 119910119894) = 119891(119909119895 119910119895) if 119894 = 119895 and 0 le119894 119895 le 119861 minus 1According to the above 119888119861 and Φ(119909 119910) can be calculated
with different B-ary For APPM proposed by Hong [10] 119888119861corresponding to B-ary is listed in Table 1 Meanwhile partsof Φ(119909 119910) corresponding to B-ary are illustrated in Figure 1
Compared with DE and EMD method APPM has theflexibility to choose a better notational system for dataembedding to decrease the image distortion The selection
4 Advances in Multimedia
Input stego image 119878Output Secret dataStep 1 Divide the stego image 119878 into non overlapping pixel pairs (1199091015840119894 1199101015840119894 )Step 2 Calculate 119904119894 = 119891(1199091015840119894 1199101015840119894 ) = (1199091015840119894 + 1198881198611199101015840119894 )mod119861 where 119894 represents the i-th pixel pairStep 3 Calculate all 119904119894 and convert them to binary stream119898
Algorithm 2
(a) Lena (b) Barbara (c) Pepper (d) Boat
(e) Tiffany (f) Baboon (g) Zelda (h) Airplane
Figure 3 The eight gray cover images
(a) Lena (b) Barbara (c) Pepper (d) Boat
(e) Tiffany (f) Baboon (g) Zelda (h) Airplane
Figure 4 The eight stego images (B=27 PSNR=45dB)
Advances in Multimedia 5
B=5B=9B=13B=17
B=21B=25B=41
46
48
50
52
54
56
58
60
62
64
PSN
R (d
B)
02 03 04 05 06 07 08 09 101Embedding Capacity ()
Figure 5 The relationships between embedding payload and image quality
of B-ary system is determined by the size of the coverimage C Given the size of C is MtimesN B is the minimumvalue satisfying lfloor119872 times 1198732rfloor ge |119904119861| However it needed tocalculate store and query the neighborhood set as shown inFigure 1
3 The Proposed Formula Adaptive Pixel PairMatching Method (FAPPM)
In order to solve the above shortcomings this paper putsforward a formula adaptive pixel pair matching embeddingmethod to find the stego-pixel pair without a neighborhoodset
31 Embedding Procedure In the embedding procedure fourvectors at most are produced Two vectors are calculatedwhen Dgt0 and the other two vectors are calculated whenDlt0 In Algorithm 1 i represents vectors 1 to 4 in turnFigure 2 shows the embedding process overview
Example 1 For a cover pixels pair (5 6) secret data 119904 = 8(16)and extraction function coefficient 11988816 = 6 the stego imagepixels pair (11990910158401 11990910158402) = (4 6) is obtained by using Algorithm 1
Step 1 Calculate 119891 = (5 + 6 times 6)mod 16 = 9 119896 = lceil(lceilradic119861rceil minus1)2rceil = 2Step 2 Calculate119863 = 119904 minus 119891 = minus1 As119863 lt 0119863 = 119863 + 119861 = 15is obtained
Step 3 Calculate 119899119890119909119905 1199051 = |119863|mod 11988816 = 3(1) Round 1 1199051 = 3 1199052 = 2(2) |1199051| gt 119896ampamp|1199052| gt 119896 then 119899119890119909119905 1199051 = 1199051 minus 119888119861 = minus3(3) Round 2 1199051 = minus3 1199052 = 3(4) |1199051| gt 119896ampamp|1199052| gt 119896 then 119863 = 119863 minus 119861 = minus1 119899119890119909119905 1199051 =minus(|119863|mod 11988816) = minus1
(5) Round 3 1199051 = minus1 1199052 = 0(6) |1199051| le 119896ampamp|1199052| le 119896 then return (4 6)
32 Extraction Procedure Through extraction functionsecret digits can be extracted from the stego image Thedetailed process is given in Algorithm 2
33 Overflow Problem and Solution If an overflow or under-flow problem occurs that is (1199091015840 1199101015840) lt 0 or (1199091015840 1199101015840) gt 255a nearest (11990910158401015840 11991010158401015840) should be found in the neighborhood of(119909 119910) such that 119891(11990910158401015840 11991010158401015840) = 119904119861 This can be done by solvingthe optimization problem
Minimize (119909 minus 11990910158401015840)2 + (119910 minus 11991010158401015840)2 Subject to 119891 (11990910158401015840 11991010158401015840) = 119904119861 0 le 11990910158401015840 11991010158401015840 le 255 (2)
4 Experimental Results and Analysis
41 Experimental Results The experiments are performedusing Matlab R2013a and eight 512 times 512 grayscale imagesare used as shown in Figure 3 The stego images are shown inFigure 4 where B=27
As seen from Figures 3 and 4 the difference between thecover images and the corresponding stego images is very littleand cannot be distinguished byhumanrsquos eyes It illustrated thegood imperceptibility of the proposed method
As message embedding it will introduce the distortion inthe image Peak signal-to-noise ratio (PSNR) is usually usedto measure the quality of image The definition of PSNR is asfollows
119875119878119873119877 = 10 times log10 ( 2552119872119878119864) (3)
6 Advances in Multimedia
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
HorizontalVertical
(a) FAPPM B=53 D=134
HorizontalVertical
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
(b) FDEMD B=53D=936
HorizontalVertical
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Occ
urre
nce
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
(c) FAPPM B=221 D=189
HorizontalVertical
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
0
500
1000
1500
2000
2500
3000
3500
4000
4500O
ccur
renc
e
(d) FDEMD B=221 D=1449
Figure 6 Comparison of the averaged vertical and horizontal difference histograms of FAPPM and FDEMD
whereMSE is themean square error between the cover imageand stego image it is defined as follows
119872119878119864 = 1119872 times 119873119872sum119894=0
119873sum119895=0
(119901119894119895 minus 1199011015840119894119895)2 (4)
Here the symbols 119901119894119895 and 1199011015840119894119895 represent the pixel valuesof the cover image and stego image in the position (119894 119895)respectively and 119872 and 119873 are the width and height of theoriginal image
As the proposed method can embed secret digit in anynotional system experiments are done to test the relationshipbetween embedding payload and image quality and the
results are shown in Figure 5 It can be found that the PSNR isdecreased as the embedding capacity is increased Howeverthe PSNR still achieved a high value when the embeddingcapacity reached 1
42 Comparison with OtherMethods Here EMD [8] EMD-3[9] APPM and FAPPM are compared from six aspects theembedding method the national system payload capacityPSNR and the storage space The results are listed in Table 2As seen from Table 2 FAPPM method uses a mathematicalmethod to embed secret data and it does not need any spaceto store neighbor table furthermore it does not affect thecapacity and image quality
Advances in Multimedia 7
Table 2 Comparison of results
Contents of comparison EMD[8] EMD-3[9] APPM[10] Proposed FAPPMEmbedding method Matrix and search Matrix and search table look-up Mathematic methodNotational systems of B-ary fixed fixed arbitrary arbitraryPayload (bpp) B=25 2471 2471 232 232PSNR (dB) 439 429 481 481Need the storage space Yes Yes Yes No
Figure 7 The difference of Rm and R-m for RS attack
Figure 8 The difference of Sm and S-m for RS attack
43 Analysis of the Security Anti-steganalysis is one of themost important criteria to measure the performance of asteganographic method In this paper a detection methodbased on histogramdifferential statistics analysis proposed byZhao [18] is used to test the security of the FAPPM methodNormally in an image with no hidingmessage the horizontaldifference histogram ℎ and the vertical difference histogramV are coincident But when the message is embedded in apair of pixels its ℎ and V will be changed The distancebetween ℎ and V is used to construct a statistical detector
to detect the variation between histograms The distance isdefined as follows
119863 = ( 2119879sum119894=minus2119879
(ℎ (119894) minus V (119894)))12
(5)
where 119879 is a predefined threshold and 119863 represents thedifference between ℎ and VThe larger the119863 is the greaterthe difference between ℎ and V is That is the probabilitythat the image contains secret information is high Hereexperiments are done to compare the histogram variationof FAPPM and FDEMD under high payload Both FAPPMand FDEMDmethods are used to generate 100 stego imagesrespectively ℎ V and their average value are calculatedrespectively The parameters are B=53 B=211 and T=20 Allthe test images were fully embedded The experiment resultsare shown in Figure 6 It can be seen that there is almostno difference between ℎ and V for FAPPM while that forFDEMD is significant which indicates the probability thatthe successful steganalysis for FDEMD is higher than that ofthe proposed method
The RS attack method can detect LSB secret data embed-ding in grayscale or color images Each pixel block is classifiedinto the regular group 119877 the singular group 119878 and theunusable group 119880 by a flipping function and mask 119872 119877119898119878119898 and 119880119898 denote the number of 119877 119878 and 119880 respectivelyFor inverse mask -119872 119877-119898 119878-119898 and U-119898 denote thenumber of 119877 119878 and 119880 respectively When no information isembedded119877119898 -119877-masymp0 and 119878119898 -119878-masymp0TheRS attack resultsare shown in Figures 7 and 8 It can be seen that the algorithmof this paper can guarantee 119877119898 -119877-masymp0 and 119878119898 -119878-masymp0 andthe existence of secret information cannot be detected by RSsteganalysis method
5 Conclusion
This paper proposed a simple and convenient data embed-ding method based on APPM Compared with the APPMmethod it has the advantage of no needing to compute andstore the neighborhood set Compared with the FDEMDmethod the secret data of any notional system is realized bythe FAPPMmethod which makes the embedding notationalsystem selection more flexible The experimental resultsshowed that FAPPM method has high image quality andthe strong anti-steganalysis ability Our future work willbe concentrated on the use of the formula method of theadjacent three pixels as the embedding unit
8 Advances in Multimedia
Data Availability
The data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
The authors declare that they have no conflicts of interest
Acknowledgments
This work was supported in part by project supported byNational Natural Science Foundation of China (Grant no61572182 no 61370225) and project supported by HunanProvincial Natural Science Foundation of China (Grant no15JJ2007)
References
[1] J Fridrich Steganography in Digital Media Principles Algo-rithms and Applications Cambridge University Press NewYork NY USA 2009
[2] A Cheddad J Condell K Curran and P Mc Kevitt ldquoDigitalimage steganography survey and analysis of current methodsrdquoSignal Processing vol 90 no 3 pp 727ndash752 2010
[3] M Hussain A W Wahab Y I Idris A T Ho and K JungldquoImage steganography in spatial domain A surveyrdquo SignalProcessing Image Communication vol 65 pp 46ndash66 2018
[4] N Provos and P Honeyman ldquoHide and seek an introduction tosteganographyrdquo IEEE Security amp Privacy vol 1 no 3 pp 32ndash442003
[5] J Fridrich M Goljan and R Du ldquoReliable Detection of LSBSteganography in Color and Grayscale Imagesrdquo The Workshopon Multimedia amp Security New Challenges ACM pp 22ndash282002
[6] A D Ker ldquoSteganalysis of LSB matching in grayscale imagesrdquoIEEE Signal Processing Letters vol 12 no 6 pp 441ndash444 2005
[7] C-K Chan and L M Cheng ldquoHiding data in images by simpleLSB substitutionrdquo Pattern Recognition vol 37 no 3 pp 469ndash474 2004
[8] C Kim ldquoData hiding by an improved exploiting modificationdirectionrdquoMultimedia Tools and Applications vol 69 no 3 pp569ndash584 2014
[9] X Niu M Ma R Tang and Z Yin ldquoImage steganography viafully exploiting modification directionrdquo International Journal ofSecurity and Its Applications vol 9 no 5 pp 243ndash254 2015
[10] W Hong and T-S Chen ldquoA novel data embedding methodusing adaptive pixel pair matchingrdquo IEEE Transactions onInformation Forensics and Security vol 7 no 1 pp 176ndash1842012
[11] J Mielikainen ldquoLSB matching revisitedrdquo IEEE Signal ProcessingLetters vol 13 no 5 pp 285ndash287 2006
[12] X Zhang and S Wang ldquoEfficient steganographic embeddingby exploiting modification directionrdquo IEEE CommunicationsLetters vol 10 no 11 pp 781ndash783 2006
[13] R Chao H Wu C Lee and Y Chu ldquoA Novel Image DataHiding Scheme with Diamond Encodingrdquo EURASIP Journal onInformation Security vol 2009 no 1 p 658047 2009
[14] W-C Kuo P-Y Lai C-C Wang and L-C Wuu ldquoA formuladiamond encoding data hiding schemerdquo Journal of Information
Hiding and Multimedia Signal Processing vol 6 no 6 pp 1167ndash1176 2015
[15] W Hong M Chen T Chen and C Huang ldquoAn efficientauthentication method for AMBTC compressed images usingadaptive pixel pair matchingrdquoMultimedia Tools amp Applicationsvol 77 no 4 pp 4677ndash4695 2018
[16] T Edwina Alias D Mathew and A Thomas ldquoSteganographicTechnique Using Secure Adaptive Pixel Pair Matching forEmbedding Multiple Data Types in Imagesrdquo in Proceedings ofthe 5th International Conference on Advances in Computing andCommunications ICACC 2015 pp 426ndash429 India September2015
[17] J Pappachan and J Baby ldquoTransformed adaptive pixel pairmatching technique for colour imagesrdquo in Proceedings of theInternational Conference on Control Instrumentation Commu-nication and Computational Technologies ICCICCT 2015 pp192ndash196 India December 2015
[18] H Zhao H Wang and M Khurram Khan ldquoStatistical analysisof several reversible data hiding algorithmsrdquo Multimedia Toolsand Applications vol 52 no 2-3 pp 277ndash290 2011
International Journal of
AerospaceEngineeringHindawiwwwhindawicom Volume 2018
RoboticsJournal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Active and Passive Electronic Components
VLSI Design
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Shock and Vibration
Hindawiwwwhindawicom Volume 2018
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawiwwwhindawicom
Volume 2018
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018
Control Scienceand Engineering
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Journal ofEngineeringVolume 2018
SensorsJournal of
Hindawiwwwhindawicom Volume 2018
International Journal of
RotatingMachinery
Hindawiwwwhindawicom Volume 2018
Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
Advances in Multimedia 3
Y
NCompute the embedding process
Are there remaining
pixelsEndGet pixelStart secret data s pair (x1x2)
Input cB and
Figure 2 The embedding process
Input A pixel pair (1199091 1199092) extraction function coefficient 119888119861 and secret data 119904Output Stego pixel pair (11990910158401 11990910158402)Step 1 Set 119891 = (1199091 + 1198881198611199092)mod119861Step 2 Set 119896 = lceil(lceilradic119861rceil minus 1)2rceilStep 3 Set119863 = 119904 minus 119891Step 4 If119863 lt 0 then119863 = 119863 + 119861Step 5 Set 119899119890119909119905 1199051 = |119863|mod 119888119861Step 6While 119894 = 1 to 4 do
Set 1199051 = 119899119890119909119905 1199051Set 1199052 = (119863 minus 1199051)119888119861If |1199051| le 119896ampamp|1199052| le 119896 then
Set 11990910158401 = 11990910158401 + 1199051Set 11990910158402 = 11990910158402 + 1199052
ElseSwitch (119894)
Case 1Set 119899119890119909119905 1199051 = 1199051 minus 119888119861
Case 2Set119863 = 119863 minus 119861Set 119899119890119909119905 1199051 = minus(|119863|mod 119888119861)
Case3Set 119899119890119909119905 1199051 = 1199051 + 119888119861
Case4Print ldquoErrorrdquo
End SwitchEnd if
End While
Algorithm 1
modified neighborhood set table and it can realize the datahiding in any notional system
2 A Review of Adaptive Pixel PairMatching (APPM)
The APPM method [10] used a pair of pixels (119909 119910) as acoordinate where an extraction function 119891119860119875119875119872(119909 119910) wasdesigned Then a neighborhood set Φ(119909 119910) of (119909 119910) wasestablished
119891119860119875119875119872 (119909 119910) = (119909 + 119888119861119910) mod119861 (1)
where 119891(119909 119910) and Φ(119909 119910) satisfied the following threeconditions
(i) In the neighborhood set Φ(119909 119910) there are exactly 119861pairs of coordinates
(ii) In the neighborhood set Φ(119909 119910) the extracted func-tion values for each coordinate aremutually exclusive
(iii) According to 119891(119909 119910) and Φ(119909 119910) a digit can beembedded in any notional system
The way to find the extraction function coefficient 119888119861and Φ(119909 119910) can be converted to find the following optimalsolution
Minimizesum119861minus1119894=0 [(119909119894minus119909)2+(119910119894minus119910)2] subject to119891(119909119894 119910119894) isin0 1 119861 minus 1 where 119891(119909119894 119910119894) = 119891(119909119895 119910119895) if 119894 = 119895 and 0 le119894 119895 le 119861 minus 1According to the above 119888119861 and Φ(119909 119910) can be calculated
with different B-ary For APPM proposed by Hong [10] 119888119861corresponding to B-ary is listed in Table 1 Meanwhile partsof Φ(119909 119910) corresponding to B-ary are illustrated in Figure 1
Compared with DE and EMD method APPM has theflexibility to choose a better notational system for dataembedding to decrease the image distortion The selection
4 Advances in Multimedia
Input stego image 119878Output Secret dataStep 1 Divide the stego image 119878 into non overlapping pixel pairs (1199091015840119894 1199101015840119894 )Step 2 Calculate 119904119894 = 119891(1199091015840119894 1199101015840119894 ) = (1199091015840119894 + 1198881198611199101015840119894 )mod119861 where 119894 represents the i-th pixel pairStep 3 Calculate all 119904119894 and convert them to binary stream119898
Algorithm 2
(a) Lena (b) Barbara (c) Pepper (d) Boat
(e) Tiffany (f) Baboon (g) Zelda (h) Airplane
Figure 3 The eight gray cover images
(a) Lena (b) Barbara (c) Pepper (d) Boat
(e) Tiffany (f) Baboon (g) Zelda (h) Airplane
Figure 4 The eight stego images (B=27 PSNR=45dB)
Advances in Multimedia 5
B=5B=9B=13B=17
B=21B=25B=41
46
48
50
52
54
56
58
60
62
64
PSN
R (d
B)
02 03 04 05 06 07 08 09 101Embedding Capacity ()
Figure 5 The relationships between embedding payload and image quality
of B-ary system is determined by the size of the coverimage C Given the size of C is MtimesN B is the minimumvalue satisfying lfloor119872 times 1198732rfloor ge |119904119861| However it needed tocalculate store and query the neighborhood set as shown inFigure 1
3 The Proposed Formula Adaptive Pixel PairMatching Method (FAPPM)
In order to solve the above shortcomings this paper putsforward a formula adaptive pixel pair matching embeddingmethod to find the stego-pixel pair without a neighborhoodset
31 Embedding Procedure In the embedding procedure fourvectors at most are produced Two vectors are calculatedwhen Dgt0 and the other two vectors are calculated whenDlt0 In Algorithm 1 i represents vectors 1 to 4 in turnFigure 2 shows the embedding process overview
Example 1 For a cover pixels pair (5 6) secret data 119904 = 8(16)and extraction function coefficient 11988816 = 6 the stego imagepixels pair (11990910158401 11990910158402) = (4 6) is obtained by using Algorithm 1
Step 1 Calculate 119891 = (5 + 6 times 6)mod 16 = 9 119896 = lceil(lceilradic119861rceil minus1)2rceil = 2Step 2 Calculate119863 = 119904 minus 119891 = minus1 As119863 lt 0119863 = 119863 + 119861 = 15is obtained
Step 3 Calculate 119899119890119909119905 1199051 = |119863|mod 11988816 = 3(1) Round 1 1199051 = 3 1199052 = 2(2) |1199051| gt 119896ampamp|1199052| gt 119896 then 119899119890119909119905 1199051 = 1199051 minus 119888119861 = minus3(3) Round 2 1199051 = minus3 1199052 = 3(4) |1199051| gt 119896ampamp|1199052| gt 119896 then 119863 = 119863 minus 119861 = minus1 119899119890119909119905 1199051 =minus(|119863|mod 11988816) = minus1
(5) Round 3 1199051 = minus1 1199052 = 0(6) |1199051| le 119896ampamp|1199052| le 119896 then return (4 6)
32 Extraction Procedure Through extraction functionsecret digits can be extracted from the stego image Thedetailed process is given in Algorithm 2
33 Overflow Problem and Solution If an overflow or under-flow problem occurs that is (1199091015840 1199101015840) lt 0 or (1199091015840 1199101015840) gt 255a nearest (11990910158401015840 11991010158401015840) should be found in the neighborhood of(119909 119910) such that 119891(11990910158401015840 11991010158401015840) = 119904119861 This can be done by solvingthe optimization problem
Minimize (119909 minus 11990910158401015840)2 + (119910 minus 11991010158401015840)2 Subject to 119891 (11990910158401015840 11991010158401015840) = 119904119861 0 le 11990910158401015840 11991010158401015840 le 255 (2)
4 Experimental Results and Analysis
41 Experimental Results The experiments are performedusing Matlab R2013a and eight 512 times 512 grayscale imagesare used as shown in Figure 3 The stego images are shown inFigure 4 where B=27
As seen from Figures 3 and 4 the difference between thecover images and the corresponding stego images is very littleand cannot be distinguished byhumanrsquos eyes It illustrated thegood imperceptibility of the proposed method
As message embedding it will introduce the distortion inthe image Peak signal-to-noise ratio (PSNR) is usually usedto measure the quality of image The definition of PSNR is asfollows
119875119878119873119877 = 10 times log10 ( 2552119872119878119864) (3)
6 Advances in Multimedia
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
HorizontalVertical
(a) FAPPM B=53 D=134
HorizontalVertical
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
(b) FDEMD B=53D=936
HorizontalVertical
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Occ
urre
nce
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
(c) FAPPM B=221 D=189
HorizontalVertical
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
0
500
1000
1500
2000
2500
3000
3500
4000
4500O
ccur
renc
e
(d) FDEMD B=221 D=1449
Figure 6 Comparison of the averaged vertical and horizontal difference histograms of FAPPM and FDEMD
whereMSE is themean square error between the cover imageand stego image it is defined as follows
119872119878119864 = 1119872 times 119873119872sum119894=0
119873sum119895=0
(119901119894119895 minus 1199011015840119894119895)2 (4)
Here the symbols 119901119894119895 and 1199011015840119894119895 represent the pixel valuesof the cover image and stego image in the position (119894 119895)respectively and 119872 and 119873 are the width and height of theoriginal image
As the proposed method can embed secret digit in anynotional system experiments are done to test the relationshipbetween embedding payload and image quality and the
results are shown in Figure 5 It can be found that the PSNR isdecreased as the embedding capacity is increased Howeverthe PSNR still achieved a high value when the embeddingcapacity reached 1
42 Comparison with OtherMethods Here EMD [8] EMD-3[9] APPM and FAPPM are compared from six aspects theembedding method the national system payload capacityPSNR and the storage space The results are listed in Table 2As seen from Table 2 FAPPM method uses a mathematicalmethod to embed secret data and it does not need any spaceto store neighbor table furthermore it does not affect thecapacity and image quality
Advances in Multimedia 7
Table 2 Comparison of results
Contents of comparison EMD[8] EMD-3[9] APPM[10] Proposed FAPPMEmbedding method Matrix and search Matrix and search table look-up Mathematic methodNotational systems of B-ary fixed fixed arbitrary arbitraryPayload (bpp) B=25 2471 2471 232 232PSNR (dB) 439 429 481 481Need the storage space Yes Yes Yes No
Figure 7 The difference of Rm and R-m for RS attack
Figure 8 The difference of Sm and S-m for RS attack
43 Analysis of the Security Anti-steganalysis is one of themost important criteria to measure the performance of asteganographic method In this paper a detection methodbased on histogramdifferential statistics analysis proposed byZhao [18] is used to test the security of the FAPPM methodNormally in an image with no hidingmessage the horizontaldifference histogram ℎ and the vertical difference histogramV are coincident But when the message is embedded in apair of pixels its ℎ and V will be changed The distancebetween ℎ and V is used to construct a statistical detector
to detect the variation between histograms The distance isdefined as follows
119863 = ( 2119879sum119894=minus2119879
(ℎ (119894) minus V (119894)))12
(5)
where 119879 is a predefined threshold and 119863 represents thedifference between ℎ and VThe larger the119863 is the greaterthe difference between ℎ and V is That is the probabilitythat the image contains secret information is high Hereexperiments are done to compare the histogram variationof FAPPM and FDEMD under high payload Both FAPPMand FDEMDmethods are used to generate 100 stego imagesrespectively ℎ V and their average value are calculatedrespectively The parameters are B=53 B=211 and T=20 Allthe test images were fully embedded The experiment resultsare shown in Figure 6 It can be seen that there is almostno difference between ℎ and V for FAPPM while that forFDEMD is significant which indicates the probability thatthe successful steganalysis for FDEMD is higher than that ofthe proposed method
The RS attack method can detect LSB secret data embed-ding in grayscale or color images Each pixel block is classifiedinto the regular group 119877 the singular group 119878 and theunusable group 119880 by a flipping function and mask 119872 119877119898119878119898 and 119880119898 denote the number of 119877 119878 and 119880 respectivelyFor inverse mask -119872 119877-119898 119878-119898 and U-119898 denote thenumber of 119877 119878 and 119880 respectively When no information isembedded119877119898 -119877-masymp0 and 119878119898 -119878-masymp0TheRS attack resultsare shown in Figures 7 and 8 It can be seen that the algorithmof this paper can guarantee 119877119898 -119877-masymp0 and 119878119898 -119878-masymp0 andthe existence of secret information cannot be detected by RSsteganalysis method
5 Conclusion
This paper proposed a simple and convenient data embed-ding method based on APPM Compared with the APPMmethod it has the advantage of no needing to compute andstore the neighborhood set Compared with the FDEMDmethod the secret data of any notional system is realized bythe FAPPMmethod which makes the embedding notationalsystem selection more flexible The experimental resultsshowed that FAPPM method has high image quality andthe strong anti-steganalysis ability Our future work willbe concentrated on the use of the formula method of theadjacent three pixels as the embedding unit
8 Advances in Multimedia
Data Availability
The data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
The authors declare that they have no conflicts of interest
Acknowledgments
This work was supported in part by project supported byNational Natural Science Foundation of China (Grant no61572182 no 61370225) and project supported by HunanProvincial Natural Science Foundation of China (Grant no15JJ2007)
References
[1] J Fridrich Steganography in Digital Media Principles Algo-rithms and Applications Cambridge University Press NewYork NY USA 2009
[2] A Cheddad J Condell K Curran and P Mc Kevitt ldquoDigitalimage steganography survey and analysis of current methodsrdquoSignal Processing vol 90 no 3 pp 727ndash752 2010
[3] M Hussain A W Wahab Y I Idris A T Ho and K JungldquoImage steganography in spatial domain A surveyrdquo SignalProcessing Image Communication vol 65 pp 46ndash66 2018
[4] N Provos and P Honeyman ldquoHide and seek an introduction tosteganographyrdquo IEEE Security amp Privacy vol 1 no 3 pp 32ndash442003
[5] J Fridrich M Goljan and R Du ldquoReliable Detection of LSBSteganography in Color and Grayscale Imagesrdquo The Workshopon Multimedia amp Security New Challenges ACM pp 22ndash282002
[6] A D Ker ldquoSteganalysis of LSB matching in grayscale imagesrdquoIEEE Signal Processing Letters vol 12 no 6 pp 441ndash444 2005
[7] C-K Chan and L M Cheng ldquoHiding data in images by simpleLSB substitutionrdquo Pattern Recognition vol 37 no 3 pp 469ndash474 2004
[8] C Kim ldquoData hiding by an improved exploiting modificationdirectionrdquoMultimedia Tools and Applications vol 69 no 3 pp569ndash584 2014
[9] X Niu M Ma R Tang and Z Yin ldquoImage steganography viafully exploiting modification directionrdquo International Journal ofSecurity and Its Applications vol 9 no 5 pp 243ndash254 2015
[10] W Hong and T-S Chen ldquoA novel data embedding methodusing adaptive pixel pair matchingrdquo IEEE Transactions onInformation Forensics and Security vol 7 no 1 pp 176ndash1842012
[11] J Mielikainen ldquoLSB matching revisitedrdquo IEEE Signal ProcessingLetters vol 13 no 5 pp 285ndash287 2006
[12] X Zhang and S Wang ldquoEfficient steganographic embeddingby exploiting modification directionrdquo IEEE CommunicationsLetters vol 10 no 11 pp 781ndash783 2006
[13] R Chao H Wu C Lee and Y Chu ldquoA Novel Image DataHiding Scheme with Diamond Encodingrdquo EURASIP Journal onInformation Security vol 2009 no 1 p 658047 2009
[14] W-C Kuo P-Y Lai C-C Wang and L-C Wuu ldquoA formuladiamond encoding data hiding schemerdquo Journal of Information
Hiding and Multimedia Signal Processing vol 6 no 6 pp 1167ndash1176 2015
[15] W Hong M Chen T Chen and C Huang ldquoAn efficientauthentication method for AMBTC compressed images usingadaptive pixel pair matchingrdquoMultimedia Tools amp Applicationsvol 77 no 4 pp 4677ndash4695 2018
[16] T Edwina Alias D Mathew and A Thomas ldquoSteganographicTechnique Using Secure Adaptive Pixel Pair Matching forEmbedding Multiple Data Types in Imagesrdquo in Proceedings ofthe 5th International Conference on Advances in Computing andCommunications ICACC 2015 pp 426ndash429 India September2015
[17] J Pappachan and J Baby ldquoTransformed adaptive pixel pairmatching technique for colour imagesrdquo in Proceedings of theInternational Conference on Control Instrumentation Commu-nication and Computational Technologies ICCICCT 2015 pp192ndash196 India December 2015
[18] H Zhao H Wang and M Khurram Khan ldquoStatistical analysisof several reversible data hiding algorithmsrdquo Multimedia Toolsand Applications vol 52 no 2-3 pp 277ndash290 2011
International Journal of
AerospaceEngineeringHindawiwwwhindawicom Volume 2018
RoboticsJournal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Active and Passive Electronic Components
VLSI Design
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Shock and Vibration
Hindawiwwwhindawicom Volume 2018
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawiwwwhindawicom
Volume 2018
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018
Control Scienceand Engineering
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Journal ofEngineeringVolume 2018
SensorsJournal of
Hindawiwwwhindawicom Volume 2018
International Journal of
RotatingMachinery
Hindawiwwwhindawicom Volume 2018
Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
4 Advances in Multimedia
Input stego image 119878Output Secret dataStep 1 Divide the stego image 119878 into non overlapping pixel pairs (1199091015840119894 1199101015840119894 )Step 2 Calculate 119904119894 = 119891(1199091015840119894 1199101015840119894 ) = (1199091015840119894 + 1198881198611199101015840119894 )mod119861 where 119894 represents the i-th pixel pairStep 3 Calculate all 119904119894 and convert them to binary stream119898
Algorithm 2
(a) Lena (b) Barbara (c) Pepper (d) Boat
(e) Tiffany (f) Baboon (g) Zelda (h) Airplane
Figure 3 The eight gray cover images
(a) Lena (b) Barbara (c) Pepper (d) Boat
(e) Tiffany (f) Baboon (g) Zelda (h) Airplane
Figure 4 The eight stego images (B=27 PSNR=45dB)
Advances in Multimedia 5
B=5B=9B=13B=17
B=21B=25B=41
46
48
50
52
54
56
58
60
62
64
PSN
R (d
B)
02 03 04 05 06 07 08 09 101Embedding Capacity ()
Figure 5 The relationships between embedding payload and image quality
of B-ary system is determined by the size of the coverimage C Given the size of C is MtimesN B is the minimumvalue satisfying lfloor119872 times 1198732rfloor ge |119904119861| However it needed tocalculate store and query the neighborhood set as shown inFigure 1
3 The Proposed Formula Adaptive Pixel PairMatching Method (FAPPM)
In order to solve the above shortcomings this paper putsforward a formula adaptive pixel pair matching embeddingmethod to find the stego-pixel pair without a neighborhoodset
31 Embedding Procedure In the embedding procedure fourvectors at most are produced Two vectors are calculatedwhen Dgt0 and the other two vectors are calculated whenDlt0 In Algorithm 1 i represents vectors 1 to 4 in turnFigure 2 shows the embedding process overview
Example 1 For a cover pixels pair (5 6) secret data 119904 = 8(16)and extraction function coefficient 11988816 = 6 the stego imagepixels pair (11990910158401 11990910158402) = (4 6) is obtained by using Algorithm 1
Step 1 Calculate 119891 = (5 + 6 times 6)mod 16 = 9 119896 = lceil(lceilradic119861rceil minus1)2rceil = 2Step 2 Calculate119863 = 119904 minus 119891 = minus1 As119863 lt 0119863 = 119863 + 119861 = 15is obtained
Step 3 Calculate 119899119890119909119905 1199051 = |119863|mod 11988816 = 3(1) Round 1 1199051 = 3 1199052 = 2(2) |1199051| gt 119896ampamp|1199052| gt 119896 then 119899119890119909119905 1199051 = 1199051 minus 119888119861 = minus3(3) Round 2 1199051 = minus3 1199052 = 3(4) |1199051| gt 119896ampamp|1199052| gt 119896 then 119863 = 119863 minus 119861 = minus1 119899119890119909119905 1199051 =minus(|119863|mod 11988816) = minus1
(5) Round 3 1199051 = minus1 1199052 = 0(6) |1199051| le 119896ampamp|1199052| le 119896 then return (4 6)
32 Extraction Procedure Through extraction functionsecret digits can be extracted from the stego image Thedetailed process is given in Algorithm 2
33 Overflow Problem and Solution If an overflow or under-flow problem occurs that is (1199091015840 1199101015840) lt 0 or (1199091015840 1199101015840) gt 255a nearest (11990910158401015840 11991010158401015840) should be found in the neighborhood of(119909 119910) such that 119891(11990910158401015840 11991010158401015840) = 119904119861 This can be done by solvingthe optimization problem
Minimize (119909 minus 11990910158401015840)2 + (119910 minus 11991010158401015840)2 Subject to 119891 (11990910158401015840 11991010158401015840) = 119904119861 0 le 11990910158401015840 11991010158401015840 le 255 (2)
4 Experimental Results and Analysis
41 Experimental Results The experiments are performedusing Matlab R2013a and eight 512 times 512 grayscale imagesare used as shown in Figure 3 The stego images are shown inFigure 4 where B=27
As seen from Figures 3 and 4 the difference between thecover images and the corresponding stego images is very littleand cannot be distinguished byhumanrsquos eyes It illustrated thegood imperceptibility of the proposed method
As message embedding it will introduce the distortion inthe image Peak signal-to-noise ratio (PSNR) is usually usedto measure the quality of image The definition of PSNR is asfollows
119875119878119873119877 = 10 times log10 ( 2552119872119878119864) (3)
6 Advances in Multimedia
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
HorizontalVertical
(a) FAPPM B=53 D=134
HorizontalVertical
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
(b) FDEMD B=53D=936
HorizontalVertical
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Occ
urre
nce
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
(c) FAPPM B=221 D=189
HorizontalVertical
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
0
500
1000
1500
2000
2500
3000
3500
4000
4500O
ccur
renc
e
(d) FDEMD B=221 D=1449
Figure 6 Comparison of the averaged vertical and horizontal difference histograms of FAPPM and FDEMD
whereMSE is themean square error between the cover imageand stego image it is defined as follows
119872119878119864 = 1119872 times 119873119872sum119894=0
119873sum119895=0
(119901119894119895 minus 1199011015840119894119895)2 (4)
Here the symbols 119901119894119895 and 1199011015840119894119895 represent the pixel valuesof the cover image and stego image in the position (119894 119895)respectively and 119872 and 119873 are the width and height of theoriginal image
As the proposed method can embed secret digit in anynotional system experiments are done to test the relationshipbetween embedding payload and image quality and the
results are shown in Figure 5 It can be found that the PSNR isdecreased as the embedding capacity is increased Howeverthe PSNR still achieved a high value when the embeddingcapacity reached 1
42 Comparison with OtherMethods Here EMD [8] EMD-3[9] APPM and FAPPM are compared from six aspects theembedding method the national system payload capacityPSNR and the storage space The results are listed in Table 2As seen from Table 2 FAPPM method uses a mathematicalmethod to embed secret data and it does not need any spaceto store neighbor table furthermore it does not affect thecapacity and image quality
Advances in Multimedia 7
Table 2 Comparison of results
Contents of comparison EMD[8] EMD-3[9] APPM[10] Proposed FAPPMEmbedding method Matrix and search Matrix and search table look-up Mathematic methodNotational systems of B-ary fixed fixed arbitrary arbitraryPayload (bpp) B=25 2471 2471 232 232PSNR (dB) 439 429 481 481Need the storage space Yes Yes Yes No
Figure 7 The difference of Rm and R-m for RS attack
Figure 8 The difference of Sm and S-m for RS attack
43 Analysis of the Security Anti-steganalysis is one of themost important criteria to measure the performance of asteganographic method In this paper a detection methodbased on histogramdifferential statistics analysis proposed byZhao [18] is used to test the security of the FAPPM methodNormally in an image with no hidingmessage the horizontaldifference histogram ℎ and the vertical difference histogramV are coincident But when the message is embedded in apair of pixels its ℎ and V will be changed The distancebetween ℎ and V is used to construct a statistical detector
to detect the variation between histograms The distance isdefined as follows
119863 = ( 2119879sum119894=minus2119879
(ℎ (119894) minus V (119894)))12
(5)
where 119879 is a predefined threshold and 119863 represents thedifference between ℎ and VThe larger the119863 is the greaterthe difference between ℎ and V is That is the probabilitythat the image contains secret information is high Hereexperiments are done to compare the histogram variationof FAPPM and FDEMD under high payload Both FAPPMand FDEMDmethods are used to generate 100 stego imagesrespectively ℎ V and their average value are calculatedrespectively The parameters are B=53 B=211 and T=20 Allthe test images were fully embedded The experiment resultsare shown in Figure 6 It can be seen that there is almostno difference between ℎ and V for FAPPM while that forFDEMD is significant which indicates the probability thatthe successful steganalysis for FDEMD is higher than that ofthe proposed method
The RS attack method can detect LSB secret data embed-ding in grayscale or color images Each pixel block is classifiedinto the regular group 119877 the singular group 119878 and theunusable group 119880 by a flipping function and mask 119872 119877119898119878119898 and 119880119898 denote the number of 119877 119878 and 119880 respectivelyFor inverse mask -119872 119877-119898 119878-119898 and U-119898 denote thenumber of 119877 119878 and 119880 respectively When no information isembedded119877119898 -119877-masymp0 and 119878119898 -119878-masymp0TheRS attack resultsare shown in Figures 7 and 8 It can be seen that the algorithmof this paper can guarantee 119877119898 -119877-masymp0 and 119878119898 -119878-masymp0 andthe existence of secret information cannot be detected by RSsteganalysis method
5 Conclusion
This paper proposed a simple and convenient data embed-ding method based on APPM Compared with the APPMmethod it has the advantage of no needing to compute andstore the neighborhood set Compared with the FDEMDmethod the secret data of any notional system is realized bythe FAPPMmethod which makes the embedding notationalsystem selection more flexible The experimental resultsshowed that FAPPM method has high image quality andthe strong anti-steganalysis ability Our future work willbe concentrated on the use of the formula method of theadjacent three pixels as the embedding unit
8 Advances in Multimedia
Data Availability
The data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
The authors declare that they have no conflicts of interest
Acknowledgments
This work was supported in part by project supported byNational Natural Science Foundation of China (Grant no61572182 no 61370225) and project supported by HunanProvincial Natural Science Foundation of China (Grant no15JJ2007)
References
[1] J Fridrich Steganography in Digital Media Principles Algo-rithms and Applications Cambridge University Press NewYork NY USA 2009
[2] A Cheddad J Condell K Curran and P Mc Kevitt ldquoDigitalimage steganography survey and analysis of current methodsrdquoSignal Processing vol 90 no 3 pp 727ndash752 2010
[3] M Hussain A W Wahab Y I Idris A T Ho and K JungldquoImage steganography in spatial domain A surveyrdquo SignalProcessing Image Communication vol 65 pp 46ndash66 2018
[4] N Provos and P Honeyman ldquoHide and seek an introduction tosteganographyrdquo IEEE Security amp Privacy vol 1 no 3 pp 32ndash442003
[5] J Fridrich M Goljan and R Du ldquoReliable Detection of LSBSteganography in Color and Grayscale Imagesrdquo The Workshopon Multimedia amp Security New Challenges ACM pp 22ndash282002
[6] A D Ker ldquoSteganalysis of LSB matching in grayscale imagesrdquoIEEE Signal Processing Letters vol 12 no 6 pp 441ndash444 2005
[7] C-K Chan and L M Cheng ldquoHiding data in images by simpleLSB substitutionrdquo Pattern Recognition vol 37 no 3 pp 469ndash474 2004
[8] C Kim ldquoData hiding by an improved exploiting modificationdirectionrdquoMultimedia Tools and Applications vol 69 no 3 pp569ndash584 2014
[9] X Niu M Ma R Tang and Z Yin ldquoImage steganography viafully exploiting modification directionrdquo International Journal ofSecurity and Its Applications vol 9 no 5 pp 243ndash254 2015
[10] W Hong and T-S Chen ldquoA novel data embedding methodusing adaptive pixel pair matchingrdquo IEEE Transactions onInformation Forensics and Security vol 7 no 1 pp 176ndash1842012
[11] J Mielikainen ldquoLSB matching revisitedrdquo IEEE Signal ProcessingLetters vol 13 no 5 pp 285ndash287 2006
[12] X Zhang and S Wang ldquoEfficient steganographic embeddingby exploiting modification directionrdquo IEEE CommunicationsLetters vol 10 no 11 pp 781ndash783 2006
[13] R Chao H Wu C Lee and Y Chu ldquoA Novel Image DataHiding Scheme with Diamond Encodingrdquo EURASIP Journal onInformation Security vol 2009 no 1 p 658047 2009
[14] W-C Kuo P-Y Lai C-C Wang and L-C Wuu ldquoA formuladiamond encoding data hiding schemerdquo Journal of Information
Hiding and Multimedia Signal Processing vol 6 no 6 pp 1167ndash1176 2015
[15] W Hong M Chen T Chen and C Huang ldquoAn efficientauthentication method for AMBTC compressed images usingadaptive pixel pair matchingrdquoMultimedia Tools amp Applicationsvol 77 no 4 pp 4677ndash4695 2018
[16] T Edwina Alias D Mathew and A Thomas ldquoSteganographicTechnique Using Secure Adaptive Pixel Pair Matching forEmbedding Multiple Data Types in Imagesrdquo in Proceedings ofthe 5th International Conference on Advances in Computing andCommunications ICACC 2015 pp 426ndash429 India September2015
[17] J Pappachan and J Baby ldquoTransformed adaptive pixel pairmatching technique for colour imagesrdquo in Proceedings of theInternational Conference on Control Instrumentation Commu-nication and Computational Technologies ICCICCT 2015 pp192ndash196 India December 2015
[18] H Zhao H Wang and M Khurram Khan ldquoStatistical analysisof several reversible data hiding algorithmsrdquo Multimedia Toolsand Applications vol 52 no 2-3 pp 277ndash290 2011
International Journal of
AerospaceEngineeringHindawiwwwhindawicom Volume 2018
RoboticsJournal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Active and Passive Electronic Components
VLSI Design
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Shock and Vibration
Hindawiwwwhindawicom Volume 2018
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawiwwwhindawicom
Volume 2018
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018
Control Scienceand Engineering
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Journal ofEngineeringVolume 2018
SensorsJournal of
Hindawiwwwhindawicom Volume 2018
International Journal of
RotatingMachinery
Hindawiwwwhindawicom Volume 2018
Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
Advances in Multimedia 5
B=5B=9B=13B=17
B=21B=25B=41
46
48
50
52
54
56
58
60
62
64
PSN
R (d
B)
02 03 04 05 06 07 08 09 101Embedding Capacity ()
Figure 5 The relationships between embedding payload and image quality
of B-ary system is determined by the size of the coverimage C Given the size of C is MtimesN B is the minimumvalue satisfying lfloor119872 times 1198732rfloor ge |119904119861| However it needed tocalculate store and query the neighborhood set as shown inFigure 1
3 The Proposed Formula Adaptive Pixel PairMatching Method (FAPPM)
In order to solve the above shortcomings this paper putsforward a formula adaptive pixel pair matching embeddingmethod to find the stego-pixel pair without a neighborhoodset
31 Embedding Procedure In the embedding procedure fourvectors at most are produced Two vectors are calculatedwhen Dgt0 and the other two vectors are calculated whenDlt0 In Algorithm 1 i represents vectors 1 to 4 in turnFigure 2 shows the embedding process overview
Example 1 For a cover pixels pair (5 6) secret data 119904 = 8(16)and extraction function coefficient 11988816 = 6 the stego imagepixels pair (11990910158401 11990910158402) = (4 6) is obtained by using Algorithm 1
Step 1 Calculate 119891 = (5 + 6 times 6)mod 16 = 9 119896 = lceil(lceilradic119861rceil minus1)2rceil = 2Step 2 Calculate119863 = 119904 minus 119891 = minus1 As119863 lt 0119863 = 119863 + 119861 = 15is obtained
Step 3 Calculate 119899119890119909119905 1199051 = |119863|mod 11988816 = 3(1) Round 1 1199051 = 3 1199052 = 2(2) |1199051| gt 119896ampamp|1199052| gt 119896 then 119899119890119909119905 1199051 = 1199051 minus 119888119861 = minus3(3) Round 2 1199051 = minus3 1199052 = 3(4) |1199051| gt 119896ampamp|1199052| gt 119896 then 119863 = 119863 minus 119861 = minus1 119899119890119909119905 1199051 =minus(|119863|mod 11988816) = minus1
(5) Round 3 1199051 = minus1 1199052 = 0(6) |1199051| le 119896ampamp|1199052| le 119896 then return (4 6)
32 Extraction Procedure Through extraction functionsecret digits can be extracted from the stego image Thedetailed process is given in Algorithm 2
33 Overflow Problem and Solution If an overflow or under-flow problem occurs that is (1199091015840 1199101015840) lt 0 or (1199091015840 1199101015840) gt 255a nearest (11990910158401015840 11991010158401015840) should be found in the neighborhood of(119909 119910) such that 119891(11990910158401015840 11991010158401015840) = 119904119861 This can be done by solvingthe optimization problem
Minimize (119909 minus 11990910158401015840)2 + (119910 minus 11991010158401015840)2 Subject to 119891 (11990910158401015840 11991010158401015840) = 119904119861 0 le 11990910158401015840 11991010158401015840 le 255 (2)
4 Experimental Results and Analysis
41 Experimental Results The experiments are performedusing Matlab R2013a and eight 512 times 512 grayscale imagesare used as shown in Figure 3 The stego images are shown inFigure 4 where B=27
As seen from Figures 3 and 4 the difference between thecover images and the corresponding stego images is very littleand cannot be distinguished byhumanrsquos eyes It illustrated thegood imperceptibility of the proposed method
As message embedding it will introduce the distortion inthe image Peak signal-to-noise ratio (PSNR) is usually usedto measure the quality of image The definition of PSNR is asfollows
119875119878119873119877 = 10 times log10 ( 2552119872119878119864) (3)
6 Advances in Multimedia
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
HorizontalVertical
(a) FAPPM B=53 D=134
HorizontalVertical
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
(b) FDEMD B=53D=936
HorizontalVertical
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Occ
urre
nce
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
(c) FAPPM B=221 D=189
HorizontalVertical
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
0
500
1000
1500
2000
2500
3000
3500
4000
4500O
ccur
renc
e
(d) FDEMD B=221 D=1449
Figure 6 Comparison of the averaged vertical and horizontal difference histograms of FAPPM and FDEMD
whereMSE is themean square error between the cover imageand stego image it is defined as follows
119872119878119864 = 1119872 times 119873119872sum119894=0
119873sum119895=0
(119901119894119895 minus 1199011015840119894119895)2 (4)
Here the symbols 119901119894119895 and 1199011015840119894119895 represent the pixel valuesof the cover image and stego image in the position (119894 119895)respectively and 119872 and 119873 are the width and height of theoriginal image
As the proposed method can embed secret digit in anynotional system experiments are done to test the relationshipbetween embedding payload and image quality and the
results are shown in Figure 5 It can be found that the PSNR isdecreased as the embedding capacity is increased Howeverthe PSNR still achieved a high value when the embeddingcapacity reached 1
42 Comparison with OtherMethods Here EMD [8] EMD-3[9] APPM and FAPPM are compared from six aspects theembedding method the national system payload capacityPSNR and the storage space The results are listed in Table 2As seen from Table 2 FAPPM method uses a mathematicalmethod to embed secret data and it does not need any spaceto store neighbor table furthermore it does not affect thecapacity and image quality
Advances in Multimedia 7
Table 2 Comparison of results
Contents of comparison EMD[8] EMD-3[9] APPM[10] Proposed FAPPMEmbedding method Matrix and search Matrix and search table look-up Mathematic methodNotational systems of B-ary fixed fixed arbitrary arbitraryPayload (bpp) B=25 2471 2471 232 232PSNR (dB) 439 429 481 481Need the storage space Yes Yes Yes No
Figure 7 The difference of Rm and R-m for RS attack
Figure 8 The difference of Sm and S-m for RS attack
43 Analysis of the Security Anti-steganalysis is one of themost important criteria to measure the performance of asteganographic method In this paper a detection methodbased on histogramdifferential statistics analysis proposed byZhao [18] is used to test the security of the FAPPM methodNormally in an image with no hidingmessage the horizontaldifference histogram ℎ and the vertical difference histogramV are coincident But when the message is embedded in apair of pixels its ℎ and V will be changed The distancebetween ℎ and V is used to construct a statistical detector
to detect the variation between histograms The distance isdefined as follows
119863 = ( 2119879sum119894=minus2119879
(ℎ (119894) minus V (119894)))12
(5)
where 119879 is a predefined threshold and 119863 represents thedifference between ℎ and VThe larger the119863 is the greaterthe difference between ℎ and V is That is the probabilitythat the image contains secret information is high Hereexperiments are done to compare the histogram variationof FAPPM and FDEMD under high payload Both FAPPMand FDEMDmethods are used to generate 100 stego imagesrespectively ℎ V and their average value are calculatedrespectively The parameters are B=53 B=211 and T=20 Allthe test images were fully embedded The experiment resultsare shown in Figure 6 It can be seen that there is almostno difference between ℎ and V for FAPPM while that forFDEMD is significant which indicates the probability thatthe successful steganalysis for FDEMD is higher than that ofthe proposed method
The RS attack method can detect LSB secret data embed-ding in grayscale or color images Each pixel block is classifiedinto the regular group 119877 the singular group 119878 and theunusable group 119880 by a flipping function and mask 119872 119877119898119878119898 and 119880119898 denote the number of 119877 119878 and 119880 respectivelyFor inverse mask -119872 119877-119898 119878-119898 and U-119898 denote thenumber of 119877 119878 and 119880 respectively When no information isembedded119877119898 -119877-masymp0 and 119878119898 -119878-masymp0TheRS attack resultsare shown in Figures 7 and 8 It can be seen that the algorithmof this paper can guarantee 119877119898 -119877-masymp0 and 119878119898 -119878-masymp0 andthe existence of secret information cannot be detected by RSsteganalysis method
5 Conclusion
This paper proposed a simple and convenient data embed-ding method based on APPM Compared with the APPMmethod it has the advantage of no needing to compute andstore the neighborhood set Compared with the FDEMDmethod the secret data of any notional system is realized bythe FAPPMmethod which makes the embedding notationalsystem selection more flexible The experimental resultsshowed that FAPPM method has high image quality andthe strong anti-steganalysis ability Our future work willbe concentrated on the use of the formula method of theadjacent three pixels as the embedding unit
8 Advances in Multimedia
Data Availability
The data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
The authors declare that they have no conflicts of interest
Acknowledgments
This work was supported in part by project supported byNational Natural Science Foundation of China (Grant no61572182 no 61370225) and project supported by HunanProvincial Natural Science Foundation of China (Grant no15JJ2007)
References
[1] J Fridrich Steganography in Digital Media Principles Algo-rithms and Applications Cambridge University Press NewYork NY USA 2009
[2] A Cheddad J Condell K Curran and P Mc Kevitt ldquoDigitalimage steganography survey and analysis of current methodsrdquoSignal Processing vol 90 no 3 pp 727ndash752 2010
[3] M Hussain A W Wahab Y I Idris A T Ho and K JungldquoImage steganography in spatial domain A surveyrdquo SignalProcessing Image Communication vol 65 pp 46ndash66 2018
[4] N Provos and P Honeyman ldquoHide and seek an introduction tosteganographyrdquo IEEE Security amp Privacy vol 1 no 3 pp 32ndash442003
[5] J Fridrich M Goljan and R Du ldquoReliable Detection of LSBSteganography in Color and Grayscale Imagesrdquo The Workshopon Multimedia amp Security New Challenges ACM pp 22ndash282002
[6] A D Ker ldquoSteganalysis of LSB matching in grayscale imagesrdquoIEEE Signal Processing Letters vol 12 no 6 pp 441ndash444 2005
[7] C-K Chan and L M Cheng ldquoHiding data in images by simpleLSB substitutionrdquo Pattern Recognition vol 37 no 3 pp 469ndash474 2004
[8] C Kim ldquoData hiding by an improved exploiting modificationdirectionrdquoMultimedia Tools and Applications vol 69 no 3 pp569ndash584 2014
[9] X Niu M Ma R Tang and Z Yin ldquoImage steganography viafully exploiting modification directionrdquo International Journal ofSecurity and Its Applications vol 9 no 5 pp 243ndash254 2015
[10] W Hong and T-S Chen ldquoA novel data embedding methodusing adaptive pixel pair matchingrdquo IEEE Transactions onInformation Forensics and Security vol 7 no 1 pp 176ndash1842012
[11] J Mielikainen ldquoLSB matching revisitedrdquo IEEE Signal ProcessingLetters vol 13 no 5 pp 285ndash287 2006
[12] X Zhang and S Wang ldquoEfficient steganographic embeddingby exploiting modification directionrdquo IEEE CommunicationsLetters vol 10 no 11 pp 781ndash783 2006
[13] R Chao H Wu C Lee and Y Chu ldquoA Novel Image DataHiding Scheme with Diamond Encodingrdquo EURASIP Journal onInformation Security vol 2009 no 1 p 658047 2009
[14] W-C Kuo P-Y Lai C-C Wang and L-C Wuu ldquoA formuladiamond encoding data hiding schemerdquo Journal of Information
Hiding and Multimedia Signal Processing vol 6 no 6 pp 1167ndash1176 2015
[15] W Hong M Chen T Chen and C Huang ldquoAn efficientauthentication method for AMBTC compressed images usingadaptive pixel pair matchingrdquoMultimedia Tools amp Applicationsvol 77 no 4 pp 4677ndash4695 2018
[16] T Edwina Alias D Mathew and A Thomas ldquoSteganographicTechnique Using Secure Adaptive Pixel Pair Matching forEmbedding Multiple Data Types in Imagesrdquo in Proceedings ofthe 5th International Conference on Advances in Computing andCommunications ICACC 2015 pp 426ndash429 India September2015
[17] J Pappachan and J Baby ldquoTransformed adaptive pixel pairmatching technique for colour imagesrdquo in Proceedings of theInternational Conference on Control Instrumentation Commu-nication and Computational Technologies ICCICCT 2015 pp192ndash196 India December 2015
[18] H Zhao H Wang and M Khurram Khan ldquoStatistical analysisof several reversible data hiding algorithmsrdquo Multimedia Toolsand Applications vol 52 no 2-3 pp 277ndash290 2011
International Journal of
AerospaceEngineeringHindawiwwwhindawicom Volume 2018
RoboticsJournal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Active and Passive Electronic Components
VLSI Design
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Shock and Vibration
Hindawiwwwhindawicom Volume 2018
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawiwwwhindawicom
Volume 2018
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018
Control Scienceand Engineering
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Journal ofEngineeringVolume 2018
SensorsJournal of
Hindawiwwwhindawicom Volume 2018
International Journal of
RotatingMachinery
Hindawiwwwhindawicom Volume 2018
Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
6 Advances in Multimedia
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
HorizontalVertical
(a) FAPPM B=53 D=134
HorizontalVertical
minus25 minus20 minus15 minus10 minus5 0 5 10 15 20 250
1000
2000
3000
4000
5000
6000
7000
(b) FDEMD B=53D=936
HorizontalVertical
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Occ
urre
nce
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
(c) FAPPM B=221 D=189
HorizontalVertical
minus20 minus15 minus10 minus5 0 5 10 15 20 25minus25Difference
0
500
1000
1500
2000
2500
3000
3500
4000
4500O
ccur
renc
e
(d) FDEMD B=221 D=1449
Figure 6 Comparison of the averaged vertical and horizontal difference histograms of FAPPM and FDEMD
whereMSE is themean square error between the cover imageand stego image it is defined as follows
119872119878119864 = 1119872 times 119873119872sum119894=0
119873sum119895=0
(119901119894119895 minus 1199011015840119894119895)2 (4)
Here the symbols 119901119894119895 and 1199011015840119894119895 represent the pixel valuesof the cover image and stego image in the position (119894 119895)respectively and 119872 and 119873 are the width and height of theoriginal image
As the proposed method can embed secret digit in anynotional system experiments are done to test the relationshipbetween embedding payload and image quality and the
results are shown in Figure 5 It can be found that the PSNR isdecreased as the embedding capacity is increased Howeverthe PSNR still achieved a high value when the embeddingcapacity reached 1
42 Comparison with OtherMethods Here EMD [8] EMD-3[9] APPM and FAPPM are compared from six aspects theembedding method the national system payload capacityPSNR and the storage space The results are listed in Table 2As seen from Table 2 FAPPM method uses a mathematicalmethod to embed secret data and it does not need any spaceto store neighbor table furthermore it does not affect thecapacity and image quality
Advances in Multimedia 7
Table 2 Comparison of results
Contents of comparison EMD[8] EMD-3[9] APPM[10] Proposed FAPPMEmbedding method Matrix and search Matrix and search table look-up Mathematic methodNotational systems of B-ary fixed fixed arbitrary arbitraryPayload (bpp) B=25 2471 2471 232 232PSNR (dB) 439 429 481 481Need the storage space Yes Yes Yes No
Figure 7 The difference of Rm and R-m for RS attack
Figure 8 The difference of Sm and S-m for RS attack
43 Analysis of the Security Anti-steganalysis is one of themost important criteria to measure the performance of asteganographic method In this paper a detection methodbased on histogramdifferential statistics analysis proposed byZhao [18] is used to test the security of the FAPPM methodNormally in an image with no hidingmessage the horizontaldifference histogram ℎ and the vertical difference histogramV are coincident But when the message is embedded in apair of pixels its ℎ and V will be changed The distancebetween ℎ and V is used to construct a statistical detector
to detect the variation between histograms The distance isdefined as follows
119863 = ( 2119879sum119894=minus2119879
(ℎ (119894) minus V (119894)))12
(5)
where 119879 is a predefined threshold and 119863 represents thedifference between ℎ and VThe larger the119863 is the greaterthe difference between ℎ and V is That is the probabilitythat the image contains secret information is high Hereexperiments are done to compare the histogram variationof FAPPM and FDEMD under high payload Both FAPPMand FDEMDmethods are used to generate 100 stego imagesrespectively ℎ V and their average value are calculatedrespectively The parameters are B=53 B=211 and T=20 Allthe test images were fully embedded The experiment resultsare shown in Figure 6 It can be seen that there is almostno difference between ℎ and V for FAPPM while that forFDEMD is significant which indicates the probability thatthe successful steganalysis for FDEMD is higher than that ofthe proposed method
The RS attack method can detect LSB secret data embed-ding in grayscale or color images Each pixel block is classifiedinto the regular group 119877 the singular group 119878 and theunusable group 119880 by a flipping function and mask 119872 119877119898119878119898 and 119880119898 denote the number of 119877 119878 and 119880 respectivelyFor inverse mask -119872 119877-119898 119878-119898 and U-119898 denote thenumber of 119877 119878 and 119880 respectively When no information isembedded119877119898 -119877-masymp0 and 119878119898 -119878-masymp0TheRS attack resultsare shown in Figures 7 and 8 It can be seen that the algorithmof this paper can guarantee 119877119898 -119877-masymp0 and 119878119898 -119878-masymp0 andthe existence of secret information cannot be detected by RSsteganalysis method
5 Conclusion
This paper proposed a simple and convenient data embed-ding method based on APPM Compared with the APPMmethod it has the advantage of no needing to compute andstore the neighborhood set Compared with the FDEMDmethod the secret data of any notional system is realized bythe FAPPMmethod which makes the embedding notationalsystem selection more flexible The experimental resultsshowed that FAPPM method has high image quality andthe strong anti-steganalysis ability Our future work willbe concentrated on the use of the formula method of theadjacent three pixels as the embedding unit
8 Advances in Multimedia
Data Availability
The data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
The authors declare that they have no conflicts of interest
Acknowledgments
This work was supported in part by project supported byNational Natural Science Foundation of China (Grant no61572182 no 61370225) and project supported by HunanProvincial Natural Science Foundation of China (Grant no15JJ2007)
References
[1] J Fridrich Steganography in Digital Media Principles Algo-rithms and Applications Cambridge University Press NewYork NY USA 2009
[2] A Cheddad J Condell K Curran and P Mc Kevitt ldquoDigitalimage steganography survey and analysis of current methodsrdquoSignal Processing vol 90 no 3 pp 727ndash752 2010
[3] M Hussain A W Wahab Y I Idris A T Ho and K JungldquoImage steganography in spatial domain A surveyrdquo SignalProcessing Image Communication vol 65 pp 46ndash66 2018
[4] N Provos and P Honeyman ldquoHide and seek an introduction tosteganographyrdquo IEEE Security amp Privacy vol 1 no 3 pp 32ndash442003
[5] J Fridrich M Goljan and R Du ldquoReliable Detection of LSBSteganography in Color and Grayscale Imagesrdquo The Workshopon Multimedia amp Security New Challenges ACM pp 22ndash282002
[6] A D Ker ldquoSteganalysis of LSB matching in grayscale imagesrdquoIEEE Signal Processing Letters vol 12 no 6 pp 441ndash444 2005
[7] C-K Chan and L M Cheng ldquoHiding data in images by simpleLSB substitutionrdquo Pattern Recognition vol 37 no 3 pp 469ndash474 2004
[8] C Kim ldquoData hiding by an improved exploiting modificationdirectionrdquoMultimedia Tools and Applications vol 69 no 3 pp569ndash584 2014
[9] X Niu M Ma R Tang and Z Yin ldquoImage steganography viafully exploiting modification directionrdquo International Journal ofSecurity and Its Applications vol 9 no 5 pp 243ndash254 2015
[10] W Hong and T-S Chen ldquoA novel data embedding methodusing adaptive pixel pair matchingrdquo IEEE Transactions onInformation Forensics and Security vol 7 no 1 pp 176ndash1842012
[11] J Mielikainen ldquoLSB matching revisitedrdquo IEEE Signal ProcessingLetters vol 13 no 5 pp 285ndash287 2006
[12] X Zhang and S Wang ldquoEfficient steganographic embeddingby exploiting modification directionrdquo IEEE CommunicationsLetters vol 10 no 11 pp 781ndash783 2006
[13] R Chao H Wu C Lee and Y Chu ldquoA Novel Image DataHiding Scheme with Diamond Encodingrdquo EURASIP Journal onInformation Security vol 2009 no 1 p 658047 2009
[14] W-C Kuo P-Y Lai C-C Wang and L-C Wuu ldquoA formuladiamond encoding data hiding schemerdquo Journal of Information
Hiding and Multimedia Signal Processing vol 6 no 6 pp 1167ndash1176 2015
[15] W Hong M Chen T Chen and C Huang ldquoAn efficientauthentication method for AMBTC compressed images usingadaptive pixel pair matchingrdquoMultimedia Tools amp Applicationsvol 77 no 4 pp 4677ndash4695 2018
[16] T Edwina Alias D Mathew and A Thomas ldquoSteganographicTechnique Using Secure Adaptive Pixel Pair Matching forEmbedding Multiple Data Types in Imagesrdquo in Proceedings ofthe 5th International Conference on Advances in Computing andCommunications ICACC 2015 pp 426ndash429 India September2015
[17] J Pappachan and J Baby ldquoTransformed adaptive pixel pairmatching technique for colour imagesrdquo in Proceedings of theInternational Conference on Control Instrumentation Commu-nication and Computational Technologies ICCICCT 2015 pp192ndash196 India December 2015
[18] H Zhao H Wang and M Khurram Khan ldquoStatistical analysisof several reversible data hiding algorithmsrdquo Multimedia Toolsand Applications vol 52 no 2-3 pp 277ndash290 2011
International Journal of
AerospaceEngineeringHindawiwwwhindawicom Volume 2018
RoboticsJournal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Active and Passive Electronic Components
VLSI Design
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Shock and Vibration
Hindawiwwwhindawicom Volume 2018
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawiwwwhindawicom
Volume 2018
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018
Control Scienceand Engineering
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Journal ofEngineeringVolume 2018
SensorsJournal of
Hindawiwwwhindawicom Volume 2018
International Journal of
RotatingMachinery
Hindawiwwwhindawicom Volume 2018
Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
Advances in Multimedia 7
Table 2 Comparison of results
Contents of comparison EMD[8] EMD-3[9] APPM[10] Proposed FAPPMEmbedding method Matrix and search Matrix and search table look-up Mathematic methodNotational systems of B-ary fixed fixed arbitrary arbitraryPayload (bpp) B=25 2471 2471 232 232PSNR (dB) 439 429 481 481Need the storage space Yes Yes Yes No
Figure 7 The difference of Rm and R-m for RS attack
Figure 8 The difference of Sm and S-m for RS attack
43 Analysis of the Security Anti-steganalysis is one of themost important criteria to measure the performance of asteganographic method In this paper a detection methodbased on histogramdifferential statistics analysis proposed byZhao [18] is used to test the security of the FAPPM methodNormally in an image with no hidingmessage the horizontaldifference histogram ℎ and the vertical difference histogramV are coincident But when the message is embedded in apair of pixels its ℎ and V will be changed The distancebetween ℎ and V is used to construct a statistical detector
to detect the variation between histograms The distance isdefined as follows
119863 = ( 2119879sum119894=minus2119879
(ℎ (119894) minus V (119894)))12
(5)
where 119879 is a predefined threshold and 119863 represents thedifference between ℎ and VThe larger the119863 is the greaterthe difference between ℎ and V is That is the probabilitythat the image contains secret information is high Hereexperiments are done to compare the histogram variationof FAPPM and FDEMD under high payload Both FAPPMand FDEMDmethods are used to generate 100 stego imagesrespectively ℎ V and their average value are calculatedrespectively The parameters are B=53 B=211 and T=20 Allthe test images were fully embedded The experiment resultsare shown in Figure 6 It can be seen that there is almostno difference between ℎ and V for FAPPM while that forFDEMD is significant which indicates the probability thatthe successful steganalysis for FDEMD is higher than that ofthe proposed method
The RS attack method can detect LSB secret data embed-ding in grayscale or color images Each pixel block is classifiedinto the regular group 119877 the singular group 119878 and theunusable group 119880 by a flipping function and mask 119872 119877119898119878119898 and 119880119898 denote the number of 119877 119878 and 119880 respectivelyFor inverse mask -119872 119877-119898 119878-119898 and U-119898 denote thenumber of 119877 119878 and 119880 respectively When no information isembedded119877119898 -119877-masymp0 and 119878119898 -119878-masymp0TheRS attack resultsare shown in Figures 7 and 8 It can be seen that the algorithmof this paper can guarantee 119877119898 -119877-masymp0 and 119878119898 -119878-masymp0 andthe existence of secret information cannot be detected by RSsteganalysis method
5 Conclusion
This paper proposed a simple and convenient data embed-ding method based on APPM Compared with the APPMmethod it has the advantage of no needing to compute andstore the neighborhood set Compared with the FDEMDmethod the secret data of any notional system is realized bythe FAPPMmethod which makes the embedding notationalsystem selection more flexible The experimental resultsshowed that FAPPM method has high image quality andthe strong anti-steganalysis ability Our future work willbe concentrated on the use of the formula method of theadjacent three pixels as the embedding unit
8 Advances in Multimedia
Data Availability
The data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
The authors declare that they have no conflicts of interest
Acknowledgments
This work was supported in part by project supported byNational Natural Science Foundation of China (Grant no61572182 no 61370225) and project supported by HunanProvincial Natural Science Foundation of China (Grant no15JJ2007)
References
[1] J Fridrich Steganography in Digital Media Principles Algo-rithms and Applications Cambridge University Press NewYork NY USA 2009
[2] A Cheddad J Condell K Curran and P Mc Kevitt ldquoDigitalimage steganography survey and analysis of current methodsrdquoSignal Processing vol 90 no 3 pp 727ndash752 2010
[3] M Hussain A W Wahab Y I Idris A T Ho and K JungldquoImage steganography in spatial domain A surveyrdquo SignalProcessing Image Communication vol 65 pp 46ndash66 2018
[4] N Provos and P Honeyman ldquoHide and seek an introduction tosteganographyrdquo IEEE Security amp Privacy vol 1 no 3 pp 32ndash442003
[5] J Fridrich M Goljan and R Du ldquoReliable Detection of LSBSteganography in Color and Grayscale Imagesrdquo The Workshopon Multimedia amp Security New Challenges ACM pp 22ndash282002
[6] A D Ker ldquoSteganalysis of LSB matching in grayscale imagesrdquoIEEE Signal Processing Letters vol 12 no 6 pp 441ndash444 2005
[7] C-K Chan and L M Cheng ldquoHiding data in images by simpleLSB substitutionrdquo Pattern Recognition vol 37 no 3 pp 469ndash474 2004
[8] C Kim ldquoData hiding by an improved exploiting modificationdirectionrdquoMultimedia Tools and Applications vol 69 no 3 pp569ndash584 2014
[9] X Niu M Ma R Tang and Z Yin ldquoImage steganography viafully exploiting modification directionrdquo International Journal ofSecurity and Its Applications vol 9 no 5 pp 243ndash254 2015
[10] W Hong and T-S Chen ldquoA novel data embedding methodusing adaptive pixel pair matchingrdquo IEEE Transactions onInformation Forensics and Security vol 7 no 1 pp 176ndash1842012
[11] J Mielikainen ldquoLSB matching revisitedrdquo IEEE Signal ProcessingLetters vol 13 no 5 pp 285ndash287 2006
[12] X Zhang and S Wang ldquoEfficient steganographic embeddingby exploiting modification directionrdquo IEEE CommunicationsLetters vol 10 no 11 pp 781ndash783 2006
[13] R Chao H Wu C Lee and Y Chu ldquoA Novel Image DataHiding Scheme with Diamond Encodingrdquo EURASIP Journal onInformation Security vol 2009 no 1 p 658047 2009
[14] W-C Kuo P-Y Lai C-C Wang and L-C Wuu ldquoA formuladiamond encoding data hiding schemerdquo Journal of Information
Hiding and Multimedia Signal Processing vol 6 no 6 pp 1167ndash1176 2015
[15] W Hong M Chen T Chen and C Huang ldquoAn efficientauthentication method for AMBTC compressed images usingadaptive pixel pair matchingrdquoMultimedia Tools amp Applicationsvol 77 no 4 pp 4677ndash4695 2018
[16] T Edwina Alias D Mathew and A Thomas ldquoSteganographicTechnique Using Secure Adaptive Pixel Pair Matching forEmbedding Multiple Data Types in Imagesrdquo in Proceedings ofthe 5th International Conference on Advances in Computing andCommunications ICACC 2015 pp 426ndash429 India September2015
[17] J Pappachan and J Baby ldquoTransformed adaptive pixel pairmatching technique for colour imagesrdquo in Proceedings of theInternational Conference on Control Instrumentation Commu-nication and Computational Technologies ICCICCT 2015 pp192ndash196 India December 2015
[18] H Zhao H Wang and M Khurram Khan ldquoStatistical analysisof several reversible data hiding algorithmsrdquo Multimedia Toolsand Applications vol 52 no 2-3 pp 277ndash290 2011
International Journal of
AerospaceEngineeringHindawiwwwhindawicom Volume 2018
RoboticsJournal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Active and Passive Electronic Components
VLSI Design
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Shock and Vibration
Hindawiwwwhindawicom Volume 2018
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawiwwwhindawicom
Volume 2018
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018
Control Scienceand Engineering
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Journal ofEngineeringVolume 2018
SensorsJournal of
Hindawiwwwhindawicom Volume 2018
International Journal of
RotatingMachinery
Hindawiwwwhindawicom Volume 2018
Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
8 Advances in Multimedia
Data Availability
The data used to support the findings of this study areavailable from the corresponding author upon request
Conflicts of Interest
The authors declare that they have no conflicts of interest
Acknowledgments
This work was supported in part by project supported byNational Natural Science Foundation of China (Grant no61572182 no 61370225) and project supported by HunanProvincial Natural Science Foundation of China (Grant no15JJ2007)
References
[1] J Fridrich Steganography in Digital Media Principles Algo-rithms and Applications Cambridge University Press NewYork NY USA 2009
[2] A Cheddad J Condell K Curran and P Mc Kevitt ldquoDigitalimage steganography survey and analysis of current methodsrdquoSignal Processing vol 90 no 3 pp 727ndash752 2010
[3] M Hussain A W Wahab Y I Idris A T Ho and K JungldquoImage steganography in spatial domain A surveyrdquo SignalProcessing Image Communication vol 65 pp 46ndash66 2018
[4] N Provos and P Honeyman ldquoHide and seek an introduction tosteganographyrdquo IEEE Security amp Privacy vol 1 no 3 pp 32ndash442003
[5] J Fridrich M Goljan and R Du ldquoReliable Detection of LSBSteganography in Color and Grayscale Imagesrdquo The Workshopon Multimedia amp Security New Challenges ACM pp 22ndash282002
[6] A D Ker ldquoSteganalysis of LSB matching in grayscale imagesrdquoIEEE Signal Processing Letters vol 12 no 6 pp 441ndash444 2005
[7] C-K Chan and L M Cheng ldquoHiding data in images by simpleLSB substitutionrdquo Pattern Recognition vol 37 no 3 pp 469ndash474 2004
[8] C Kim ldquoData hiding by an improved exploiting modificationdirectionrdquoMultimedia Tools and Applications vol 69 no 3 pp569ndash584 2014
[9] X Niu M Ma R Tang and Z Yin ldquoImage steganography viafully exploiting modification directionrdquo International Journal ofSecurity and Its Applications vol 9 no 5 pp 243ndash254 2015
[10] W Hong and T-S Chen ldquoA novel data embedding methodusing adaptive pixel pair matchingrdquo IEEE Transactions onInformation Forensics and Security vol 7 no 1 pp 176ndash1842012
[11] J Mielikainen ldquoLSB matching revisitedrdquo IEEE Signal ProcessingLetters vol 13 no 5 pp 285ndash287 2006
[12] X Zhang and S Wang ldquoEfficient steganographic embeddingby exploiting modification directionrdquo IEEE CommunicationsLetters vol 10 no 11 pp 781ndash783 2006
[13] R Chao H Wu C Lee and Y Chu ldquoA Novel Image DataHiding Scheme with Diamond Encodingrdquo EURASIP Journal onInformation Security vol 2009 no 1 p 658047 2009
[14] W-C Kuo P-Y Lai C-C Wang and L-C Wuu ldquoA formuladiamond encoding data hiding schemerdquo Journal of Information
Hiding and Multimedia Signal Processing vol 6 no 6 pp 1167ndash1176 2015
[15] W Hong M Chen T Chen and C Huang ldquoAn efficientauthentication method for AMBTC compressed images usingadaptive pixel pair matchingrdquoMultimedia Tools amp Applicationsvol 77 no 4 pp 4677ndash4695 2018
[16] T Edwina Alias D Mathew and A Thomas ldquoSteganographicTechnique Using Secure Adaptive Pixel Pair Matching forEmbedding Multiple Data Types in Imagesrdquo in Proceedings ofthe 5th International Conference on Advances in Computing andCommunications ICACC 2015 pp 426ndash429 India September2015
[17] J Pappachan and J Baby ldquoTransformed adaptive pixel pairmatching technique for colour imagesrdquo in Proceedings of theInternational Conference on Control Instrumentation Commu-nication and Computational Technologies ICCICCT 2015 pp192ndash196 India December 2015
[18] H Zhao H Wang and M Khurram Khan ldquoStatistical analysisof several reversible data hiding algorithmsrdquo Multimedia Toolsand Applications vol 52 no 2-3 pp 277ndash290 2011
International Journal of
AerospaceEngineeringHindawiwwwhindawicom Volume 2018
RoboticsJournal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Active and Passive Electronic Components
VLSI Design
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Shock and Vibration
Hindawiwwwhindawicom Volume 2018
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawiwwwhindawicom
Volume 2018
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018
Control Scienceand Engineering
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Journal ofEngineeringVolume 2018
SensorsJournal of
Hindawiwwwhindawicom Volume 2018
International Journal of
RotatingMachinery
Hindawiwwwhindawicom Volume 2018
Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
International Journal of
AerospaceEngineeringHindawiwwwhindawicom Volume 2018
RoboticsJournal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Active and Passive Electronic Components
VLSI Design
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Shock and Vibration
Hindawiwwwhindawicom Volume 2018
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawiwwwhindawicom
Volume 2018
Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom
The Scientific World Journal
Volume 2018
Control Scienceand Engineering
Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom
Journal ofEngineeringVolume 2018
SensorsJournal of
Hindawiwwwhindawicom Volume 2018
International Journal of
RotatingMachinery
Hindawiwwwhindawicom Volume 2018
Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawiwwwhindawicom Volume 2018
Hindawiwwwhindawicom Volume 2018
Navigation and Observation
International Journal of
Hindawi
wwwhindawicom Volume 2018
Advances in
Multimedia
Submit your manuscripts atwwwhindawicom
Related Documents
![W-PoseNet: Dense Correspondence Regularized Pixel Pair ... · Pixel-wise features are sparsely sampled and combined to generate pixel pair features, which produces 6D pose [Rjt] and](https://static.cupdf.com/doc/110x72/5f1f9604d751a4018357f889/w-posenet-dense-correspondence-regularized-pixel-pair-pixel-wise-features-are.jpg){kind=tracking}
