comparative study of watermarking and encryption schemes for jpeg2000 images

4

Click here to load reader

Upload: al-musanna-college-of-technologysultanate-of-oman

Post on 11-Apr-2017

28 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Comparative Study of Watermarking and Encryption Schemes for JPEG2000 Images

International Journal of Applied Engineering Research, ISSN 0973-4562 Vol. 9 No.26 (2014) pp. 9323-9326

© Research India Publications; http://www.ripublication.com/ijaer.htm

Comparative Study of Watermarking and Encryption

Schemes for JPEG2000 Images

.J.Jereesha Mary, 2S.Sebastin Antony Joe,

3Dr.C.Seldev Christopher

Abstract: The need for secured digital data is getting more intense

and hence there is a necessity for ownership verification, tamper

proof and data integrity. As a solution for these, digital

watermarking methods may be used. In this paper, three

watermarking schemes namely Spread Spectrum (SS), Scalar

Costa scheme Quantization Index Modulation (SCS_QIM),

Rational Dither Modulation (RDM) and the encryption techniques

like RC4,RC5,RC6 which are responsible for the robustness of the

above said watermarking system for JPEG2000 images are

compared. The Peak Signal to Noise Ratio (PSNR) and the Mean

Square Error (MSE) for different watermark images have been

compared and presented in this paper.

Keywords – RC4, RC5, RC6, Spread Spectrum, Scalar Costa Scheme

Quantization Index modulation, Rational Dither

Modulation,JPEG2000.

INTRODUCTION

Digital media communication and security for the digital content has seen a rapid growth in the current era. Digital Asser Management Systems (DAMS) are intended to work with compressed and encrypted data. Media contents liketext, image, audio, video, etc., can be secured using digital watermark techniques. Watermarking is a method that is used for reducing counterfeiting in which the watermark is embedded inside the message. This watermark can then be extracted or detected for variety of purposes. Image authentication with watermarking schemes was proposed by many researches in order to overcome the drawbacks of image authentication using digital signature schemes. In the latter scheme the modification of image or tampering can be identified,but the exact location of modification cannot be found [1-4]. Fridrich [5] proposed a block cipher method to prevent collage attack and Wong [6] proposed a public key fragile watermarking scheme for image authentication which was vulnerable to vector quantization attack. This was proven by Memon[7] and hence WongandMemon[8] proposed an improved model to reduce vector quantization attack. The above said methods are for JPEG compressed images.

In this paper, we present a comparative study on various watermarking schemes like spread spectrum, SCS-QIM, RDM and the encryption methods used in coordination with those watermarking schemes for JPEG2000 images. Compressed images are mainly used in digital watermarking algorithms. Watermarking algorithms are said to be effective if the embedded watermark is robust against compression. The

encryption is done in the compressed JPEG2000 image except the header and marker segment [9]. JPEG2000 images are being used, since the decompression can be made at any ratio for the same compression ratio [10]. There aremany encryption algorithms proposed to work with watermarking algorithms. Using asymmetric schemes with homomorphic property, the RSA, Elgamel, Goldwasser-Micali, Paillierhave their own drawbacks like, the ciphertext output results in loss of compression efficiency if the message size is small and if the message size is large the compression loss is reduced, but the payload capacity decreases [11-14]. To overcome this Subramanyam et al. in [15] proposed RC4 symmetric stream cipher scheme with homomorphic property to achieve copyright protection. Gayathri I.K. proposed the symmetric block cipher RC5 [16] to work with the watermarking schemes used in [15]. RC5 is more secure and robust than RC4 due to its increased number of rounds. RC6 is being used by Kukoo Anna Mathew [17] instead of RC5 due to its increased use of registers.

II. TECHNICAL BACKGROUND

The watermarking schemes used to provide robustness and the encryption schemes to secure the JPEG2000 images for copy right protection and content authentication are listed below.

JPEG2000 standard JPEG2000 is a compression standard that is based on

wavelet technology. Scalability and editability are its main advantage. It meansfor a single encoding resolution, it can be decoded at any resolution. Hence it needs computationally complex encoders and decoders. JPEG2000 supports very low and very high compression rates. It define both the file formats and code streams. In JPEG2000 compression technique, the color components of the input are transformed using irreversible color or reversiblecolor transform. The output is tiled and then wavelet transformed to an arbitrary depth using irreversible orreversibletransform. It is then quantized and coded using Embedded Block Coding with Optimized Truncation (EBCOT) scheme to get the compressed output [18].The JPEG2000 compression standard processesthe output as byte streams.

Encryption Schemes Encryption algorithms are used in co-ordination with the

watermarking schemes to make the JPEG2000 images more robust so that it provides ownership verification & content authentication.Various encryption schemes are discussed below,

RC4

Page 2: Comparative Study of Watermarking and Encryption Schemes for JPEG2000 Images

International Journal of Applied Engineering Research, ISSN 0973-4562 Vol. 9 No.26 (2014)

© Research India Publications; http://www.ripublication.com/ijaer.htm

9324

RC4 is a symmetric stream cipher with homomorphic property. RC4 generates a pseudorandom stream of bits (a keystream). As with any stream cipher, these can be used for encryption by combining them with the plaintext using bit-wise Exclusive-OR. Encryption is done in the input image byte by byte. This is shown as,

C=E(M,Ks) C =(mi + ki) mod 255 i = 0,1,2. . . .L-1 WhereM - Input image C -cipher text E - Encryption algorithm Ks -key Key generation is an important part in RC4 in which a secret

valueSvsharable by both the encryption and decryption side is applied to RC4 which in turn generates the key stream ks[19].

RC5 RC5 is a homomorphic, symmetric-key block cipher which

has a variable block size (32, 64 or 128 bits), key size (0 to 2040 bits) and number of rounds (0 to 255).RC5 uses two registers and modular addition. Key schedule is a complex part in RC5 [20].

A = A + S[0]; B = B + S[1]; for i = 1 to r do A= ((A Xor B) <<< B) + S[ 2 * i ] B = ((B Xor A) <<< A) + S[ 2 * i + 1] Where A, B - two w-bit words of the input in input registers S - Expanded key table RC6 RC6 is a symmetric keyblock cipher derived from RC5. Like

RC5, it can be parameterized to support a wide variety of word-lengths, key sizes and number of rounds. But RC6 does include the use offour working registers instead of twoand uses an extra integer multiplication operation in order to make the rotation dependent on every bit in a word, and not just the least significant few bits.The encryption process is composed of three stages viz., pre-whitening, inner loop of rounds, and post-whitening. Plaintext revealing part of the input to the first round and the ciphertext revealing part of the input to the last round of encryption are removed bypre-whitening and post-whitening.The operations used in RC6 are defined as followings [21].

A+B integer addition modulo 2w A-B integer subtraction modulo 2w A⊕B bitwise exclusive-or of w-bit words A x B integer multiplication modulo 2w A⋘B rotation of the w-bit word A to the left by theamount

given by the least significant log w bits ofB A⋙B rotation of the w-bit word A to the right by theamount

given by the least significant log w bits ofB (A,B,C,D)=(B,C,D,A) parallel assignment. Watermarking Algorithms

Watermarking algorithms are used to embed the watermark inside the input image. The watermarking algorithm should be robust for ownership verification and fragile for content authentication.Image watermarking process should be one without having any visual artifacts between the original and the watermarked image. Few watermarking schemes are being discussed below.

Spread Spectrum (SS) Spread spectrum system has thetransmitted signal spread

over a wide frequency band.The codes are unique to every user and have low cross correlation value. The main advantage of spread spectrum is it blocks purposeful and accidentalinterference. Since it uses pseudo random code sequence, the privacy of the code is maintained [22]. The watermarked vector may be expressed as

Wv= S + W Where Wv- Watermarked vector S- Signal vector W- Watermark vector Both S and W are in the same vector space R Scalar costa scheme(SCS) Scalar costa scheme is a technique for embedding that

belongs to quantization based methods and it uses scalar approach and reception function [23]. In scalar costa scheme the Lx – dimensional codebook of costa scheme is replaced by a product codebook that contains scalar quantizer of uniform step

size . The watermarked data is given as Wm = M +W=M+ α (Qi(M)-M) Where, Wm- Watermarked sequence. M - Message W - Watermark sequence α- Scalar factor / Distortiioncompensation parameter Qi(M) - Quantization value of M using step size D. Rational Dither Modulation (RDM) RDM is a modification of Dither Modulation (DM) which

overcomes the vulnerability of scaling attack. It does not focus on the step-size of the quantizer and hence avoids the usage of pilot-sequence. It is used for carrying outscalar operation and hence avoids the construction of spherical codes. RDMuses gain-invariant adaptive quantization step-size at both embedder and decoder. This causes the watermarked signal being asymptotically stationary [25]. The embedding rule is given as follows

where yk- vector with„L‟ past samples of vector „y‟ g- isa function, that satisfies with

-Quantiers

III. SECURITY OF CRYPTOSYSTEM

RC4

Page 3: Comparative Study of Watermarking and Encryption Schemes for JPEG2000 Images

International Journal of Applied Engineering Research, ISSN 0973-4562 Vol. 9 No.26 (2014)

© Research India Publications; http://www.ripublication.com/ijaer.htm

9325

RC4 crypto system has the advantage of being faster and it is suitable for streaming application [19].[26] proved that relatively low sampling resistancein RC4 will increase the probability of data tradeoff attacks based on the key scheduling algorithm. Also the RC4 algorithm is weak in common mode of operation and is commonly used in Wired Equivalent Privacy Protocol (WEP).

RC5 RC5 algorithm is simple. The data dependent rotations and

the complexity in key scheduling are its advantages [27]. RC5-32algorithm (32-bit blocks) is susceptible to known plain text attack which is linear.From[28]it is shown that RC5-32 has a factor of approximately 26 to attack an extra round and RC5-64 has around 28.RC5-64 algorithm (64-bit blocks)is prone to differential attack, when the chosen plain text is 244. But it overcomes the attack with 18 – 20 rounds.

RC6 RC6 is also susceptible to differential linear attack as that of

its predecessor RC5[28]. Linear attack is possible with 16 rounds with text size 2119. With less than 15 rounds it undergoes statistical attack [29]. RC6 is affected by X2 attack when the number of rounds is more than 16[30]

IV. ROBUSTNESS OF EMBEDDING ALGORITHM

Spread Spectrum An attack on spread spectrum watermarks based on

estimation of the watermarks is explained in [31]. Various attacks that spread spectrum watermarking prone to are simple attacks, detection disabling attacks, ambiguity attack, removal attack etc.

Scalar Costa Scheme with QIM Under AWGN attack the performance of Scalar Costa

Scheme is high compared to spread spectrum. Sospread spectrum is of no use even when the watermark must survive high noise level [24]. For high noise levels independent of the host statistics and the document to watermark ratio the achievable embedding rate of scalar costa scheme is smaller than spread spectrum. When the watermark to noise ratio increases, the scalar costa scheme outperforms spread spectrum [24].

Rational Dither Modulation Rational Dither Modulation has higher capacity than spread

spectrum and scalar costa scheme. It has benefits from the gains afforded by distortion compensation and channel coding. Moderately it is robust to Invariant value metric scaling attack [25]. The RDM as in [25] poorly models the human perceptual system and a function that models the properties of perception should be considered [32]

V. CONCLUSION

The above comparative study discloses that a number of watermarking and encryption schemes can be used for different watermarking applications. The watermarking schemes compared are robust and they can be used for content authentication. The encryption algorithms which are explained

in this paper can be used in addition to the watermarking schemes and can be applied for copy right protection. The advantages and drawbacks the major water marking / encryption schemes are being discussed here. The watermarking schemes can all be together used with a single encrypion algorithm. The research work can be extended to find a suitable encryption scheme and watermarking algorithm to be put together to give a robust watermarking for JPEG2000 images.

VI. REFERENCES

[1] Eggers JJ, Girod B. Blind watermarking applied to image authentication. In:Proceedings of IEEE international conference on acoustics, speech and signalprocessing. 2001. p. 1977–80.

[2] Kundur D, Hatzinakos D. Digital watermarking for telltale tamper proofing andauthentication. In: Proceedings of the IEEEspecial issue on identification andprotection of multimedia information. 1999. p. 1167–80.

[3] Lin ET, Podilchuk CI, Delp EJ. Detection of imagealterations using semi-fragilewatermarks. In: Proceedings of SPIEconference on security and watermarkingof multimedia contents II. 2000. p. 152–63.

[4] Wang MS, Chen WC. A majority-voting based watermarking scheme forcolor image tamper detection and recovery. Computer Standards & Interfaces2007;29:561–70.

[5] Fridrich J, Goljan M, Baldoza AC. New fragile authentication watermark forimages. In: Proceeding of IEEE international conference on image processing,vol. 1. 2000. p. 446–9.

[6] Memon N, Shende S, Wong P. On the security of the Yeung–Mintzer authenticationwatermark. In: Proceedings of the IS&T PICS symposium. 1999. p. 301–6.

[7] Holliman M, Memon N. Counterfeiting attacks on oblivious block-wiseindependent invisible watermarking schemes. IEEE Transactions on Image Processing2000;9(3):432–41.

[8] Wong PW, Memon N. “Secret and public key image watermarking schemes forimage authentication and ownership verification”. IEEE Transactions on Image Processing 2001;10(10):1593–601.

[9] H. Wu and D.Ma, “Efficient and secure encryption schemes for JPEG2000,” in Proc. IEEE Int. Conf. Acoustics, Speech and Signal Processing, 2004, vol. 5, pp. 869–872.

[10] M. Rabbani and R. Joshi, “An overview of the JPEG 2000 still image compression standard,” Signal Process.: Image Commun., vol. 17, no.1, pp. 3–48, 2002.

[11] R. Rivest, A. Shamir, and L. Adleman, “A method for obtaining digital signatures and public-key cryptosystems,” Commun. ACM, vol. 21, no.2, pp. 120–126, 1978.

[12] S. Goldwasser and S. Micali, “Probabilistic encryption,” J. Comput.Syst. Sci., vol. 28, no. 2, pp. 270–299, 1984.

[13] T. ElGamal, “A public key cryptosystem and a signature scheme based on discrete logarithms,” IEEE Trans. Inf. Theory, vol. 31, no. 4, pp.469–472, Jul. 1985.

[14] P. Paillier, “Public-key cryptosystems based on composite degree residuosity classes,” Lecture Notes in Computer Science, pp. 223–238,1999

[15] [15] A. Subramanyam, S. Emmanuel, and M. Kankanhalli, “Compressed encrypted domain JPEG2000 image watermarking,” in Proc. IEEE Int. Conf. Multimedia and Expo, 2010, pp. 1315–1320.

[16] [16]Gayathri I. K, “Digital Watermarking Using RC5 Encryption On JPEG2000 Images” International Journal of Engineering Research & Technology (IJERT) Vol. 2 Issue 7, July – 2013

[17] Mathew, Kukoo Anna. "Watermarking of JPEG2000 Compressed Images with Improved Encryption." International Journal of Computer Applications Technology and Research 2, no. 3: 245-249.

Page 4: Comparative Study of Watermarking and Encryption Schemes for JPEG2000 Images

International Journal of Applied Engineering Research, ISSN 0973-4562 Vol. 9 No.26 (2014)

© Research India Publications; http://www.ripublication.com/ijaer.htm

9326

[18] C. Christopoulos, A. Skodras, T. Ebrahimi, “The JPEG2000 Still Image Coding System : An Overview”, IEEE Transactions on Consumer Electronics, Vol. 46, No. 4, pp.1103-1127, november 2000.

[19] AllamMousa and Ahmad Hamad, "Evaluation of the RC4 Algorithm for Data Encryption", International Journal of Computer Science & Applications Vol. 3, No. 2 , June 2006, pp 44-56. (http://www. tmrfindia. org/ijcsa/V3I24. pdf).

[20] H. Eldean Ahmed, H. M. Kalash, and O. S. Faragallah, "Implementation of RC5 block cipher algorithm for image cryptosystem," International Journal of Information Technology, vol. 3, no. 4, pp. 245-250, 2004.

[21] Rivest R.L., Robshaw M.J.B., Sidney R.,YinY.L., (1998) ”The RC6block cipher.” AES-The First AdvancedEncryptionStandardCandidateConference, Conference Proceeings. http://people.csail.mit.edu/rivest/Rc6.pdf

[22] F. Hartung, J.K. Su, and B. Girod, “Spread spectrum watermarking: Malicious attacks and counterattacks,” in Proc. SPIE Security and Watermarking of Multimedia Contents 99, pp. 147–158, 1999.

[23] J. Eggers, R. Bauml, R. Tzschoppe, and B. Girod, “Scalar costa scheme for information embedding”, IEEE Trans. Signal Process., vol. 51, no. 4, pp. 1003–1019, Apr. 2003.

[24] L. Perez-Freire, F. Perez-Gonzalez, and S. Voloshynovskiy, "Revealing the true achievable rates of scalar Costa scheme," inProc. IEEE International Workshop on Multimedia Signal Processing (MMSP), Siena, Italy, 2004.

[25] Perez-Gonzalez, F., Mosquera, C., Barni, M., and Abrardo, A.: „Rational dither modulation: a high-rate data-hiding method invariant to gain attacks‟. IEEE Trans. Signal Process., 2005, 53, pp. 3960–3975.

[26] S. Fluhrer, I. Mantin, and A. Shamir, “Weaknesses in the key scheduling algorithm of RC4,” in Selected Areas in Cryptography, 2001.

[27] B.S. Kaliski, Y.L. Yin, “On the Security of the RC5 Encryption Algorithm,"RSA Laboratories Technical Report TR-602, Version 1.0, September 1998.

[28] Johan Borst, Bart Preneel, JoosVandewalle K.U. Leuven, “Linear Cryptanalysis of RC5 and RC6”, FSE 1999, Lecture Notes in ComputerScience, Springer-Verlag, pp. 16-30, 1999.

[29] H. Gilbert, H.Handschuh, A. Joux, and S.Vaudenay, “Statistical Attack on RC6”, FSE2000, LNCS 1978,pp.64-74, Springer-Verlag,2000

[30] A. Miyaji and T. Takano “Evaluation of the security of RC6 against the x2-attack”, IEICE Trans., Fundamentals. vol. E90-A, No.1(2007), 22-28.

[31] G. Langelaar, R. Lagendijk, and J. Biemond, “Removing spatial spread spectrum watermarks by non-linear_ltering," in Proceedings European Signal Processing Conference (EUSIPCO 98), Rhodes, Greece, September 1998.

[32] Cox I. J., LI Qiao. “Rational dither modulationwatermarking using a perceptual model”, 2005 IEEE 7thWorkshop on Multimedia Signal Processing, Oct, 2005