DFT BASED INDIVIDUAL EXTRACTION OF STEGANOGRAPHIC .

5m ago
33 Views
0 Downloads
430.90 KB
5 Pages
Transcription

IJRET: International Journal of Research in Engineering and TechnologyeISSN: 2319-1163 pISSN: 2321-7308DFT BASED INDIVIDUAL EXTRACTION OF STEGANOGRAPHICCOMPRESSION OF IMAGESC. Rengarajaswamy11Assistant Professor, Department of Electronics and Communication Engineering, M.A.M School of Engineering, Trichy,Tamil NaduAbstractA Novel scheme of Separable Extraction of concealed data and compressed image is proposed adopting a transform basedcompression over the encrypted data. Reversible data hiding is branched under Steganography. In this reversible data hiding the artof concealing a secret data over the cover image is practiced. More than that, the main ingredients that play the key role areencryption key and the data hiding key, the compression is performed in between the two processes i.e., encryption and data hiding, tocreate a spare space to accommodate the secret data. The main aim of this proposed scheme is to limit the distortion and to enhance awide space for concealing of the secret data i.e., to increase the embedding rate. The compression here classified into lossy andlossless. Lossy compression best suits for image and audio (telephone conversation, etc.) whereas lossless compression needed in thetext and videos. Hence to perform the lossless compression over a stegomized image and improve its secrecy, DFT is performed afterEncryption.Keywords: DFT, Data Embedding, RDH, Separable RDH, ---1. INTRODUCTIONIn the world of communication, security plays a vital role andowns a major regime on its stack. The word communicationnow a day’s does not travel alone, but hand in hand withsecurity aspects. Thus security turns to be the key to unlock acommunication box. Coming to security, which is branchedinto cryptography, information hiding and watermarking etc,provides better role as they take part. As the technology keepson changing its face with advanced features, there is need toget updated to it in its long run towards the betterment offuture. Usually the happening is that when a new algorithm isproduced or an existing algorithm is revised, intruders orhackers break the algorithm. So it is a must to developalgorithms more efficient and be stable and unbreakable tomost extent. Normally, the network security is branched intocryptography and information hiding. Information hidingcontains Steganography and watermarking which may bedated back old compared to cryptography. In cryptography,scramble of data’s takes place at the transmitter andunscrambling them gives the exact input at the receiversection. Thus for scrambling and unscrambling denotedtechnically as encryption and decryption, a key is used. Thusto encrypt and decrypt same key or different keys may beused. Further extending its branches into symmetric(conventional) and asymmetric (public key) encryption. Herein the former encryption, same key is used both at thetransmitter and receiver. Bit in the later case, different key ishandled. Coming to Steganography, which is art of concealingof data into other. It may be dated past, but again heads tomore secure transmission. In order to improve the securitylevel, combination of various techniques is handled. One suchtry is the Steganography over cryptography. Thus both thetechniques provide better authentication and integrity amongthe users. One such technique chosen under Steganography isRDH (Reversible Data Hiding). In RDH, major ingredients arethe cover data and the secret data. Cover data is which thesecret data to be hidden, like the letter enclosed in anenvelope. The other is secret or additional data. Here theoverall data enclosing the cover data and secret data is calledthe marked cover. In RDH more importance is given to thecover data, such that the user information is inscribed over it.There is a lot application employing RDH mainly in the caseof military communication, patient details, in case ofemergency over country, etc. The below diagram illustratesthe basic block of how RDH is processed through. Thusembedding secret data to the cover data at transmitter and thereverse process explains abut extracting the secret data as wellas the cover data without any error, if error data not recovered.The later part of this paper is lapsed as follows. Section IIcomprises related works describing RDH and its separableextension. Section III describes the proposed work handlingstream cipher encryption and DFT compression andembedding. Section IV gives us the exact view of theproposed scheme as the experimental results using ImageProcessing Tool. Section V summarizes the whole content to ashort and sweet manner. Further, in section VI the referencesare listed.Volume: 03 Issue: 02 Feb-2014, Available @ http://www.ijret.org199

IJRET: International Journal of Research in Engineering and TechnologyTransmittereISSN: 2319-1163 pISSN: processSecretImageSecretImageFig 1 Basic Block diagram of RDH2. RELATED WORKS2.1 Embedding in LSBA digital image is basically divided into pixels. These pixelvalues may vary for each type of images. For a Binary image,the pixel value is 1, for gray scale and color the range of pixelis 8-bit and 24-bit. So to carry the LSB embedding process,mostly gray scale and color images are preferred. Here in LSBonly the least bits of a pixel is altered or cleared to embed thesecret or additional data. If any change is made in the MSBbits, they change the whole nature of the Image. Here thePayload Data (secret Data) is embedded in such a way that canbe reconstructed only at the receiving end. The commonprocedure that is carried out to embed a payload data is givenby the formula [1],Si Ci Ci mod 2k mithe transmitter side the data is embedded into the cover imagein the sparse space already created to hide data. Here theadditional data hidden in done using the data-hiding key, sothat the same key is used at the receiver to recover the secretdata. As soon as the compression and embedding process thedata is passed to the receiving end along with the data-hidingkey. In the receiver end, the extraction process is carried outusing the same data-hiding key.During this process if any change is done in the transmitteddata, it causes some error in the transmitted data. Thus if anyerror occurs either in the cover image r the secret data, theoriginal data cannot be retrieved. This is the special propertyof using RDH which provides additional authentication andintegrity.2.3 RDH over Encrypted ImagesThus LSB is the most basic method and used in common forcreating the sparse space. Thus the sparse space created isuseful for hiding the additional payload data. This makes thework easier. Hence at the extraction part it is easier to retrievethe original message as well as the secret data with low dataloss.In this section, RDH is performed over an encrypted Image.Thus it carries three steps and two keys which is same for boththe transmitter and the receiver. The first is encryption usingencryption key and sparse space creation and third is the dataembedding using data-hiding key. The receiver is done in anorderly process i.e., decryption using the encryption key torecover the cover image and then applying the data-hiding keyto recover the secret data. If any key is applied without thespecified manner the cover image and the secret data cannotbe recovered, which is a unique property of RDH [2].2.2 Reversible Data Hiding2.4 Separable RDH over Encrypted ImagesReversible data hiding (RDH) is a technique which is entirelydifferent from other substitution techniques. In RDH, the mainaim is to recover the data without any error. If there is anyerror either in the cover data or the secret explained asfollows. The above diagram shows us how the process issuccessfully carried out using RDH. The RDH consists of twosteps each carried at the transmitter and the receiver end. AtIn the earlier method proposed [2], at the receiver enddecryption is performed first in order to extract the coverimage and then applying the data-hiding key to recover thesecret data, which is non-separable in manner. In the case ofseparable RDH, either of the key can be used irrespective ofthe other to recover the respective data’s independently [3]Here Si - stego image, Ci - the cover image, k - the embeddingparameter and mi - the payload.Volume: 03 Issue: 02 Feb-2014, Available @ http://www.ijret.org200

IJRET: International Journal of Research in Engineering and TechnologyeISSN: 2319-1163 pISSN: 2321-7308was proposed. This proves to be better flexible among the userhandling different keys.Original Cover ImageCover ImageEncryptionKeyEncryptionKeyDecryptionStream ressed ImageSecretImageInformationHidingData ExtractionStego ImageFig 2 Proposed RDH using DFT3. PROPOSED SCHEME3.1 Encryption Using Stream CipherIn the encryption part, it is a must to choose what type ofencryption is adopted and which is best suitable. Streamcipher encryption is chosen despite its high security andflexibility. In stream cipher, the encryption in carried as bit bybit format, whereas in the block cipher, an entire block issubjected to encryption. In stream cipher, a raw data is XORed with the randomly generated key by the pseudorandomgenerator. The pseudo random generated key is used both atthe transmitter and the receiver. Thus the same key is used todecrypt the data at the receiver end, hence the name symmetrickey encryption.Di,j,u di,j,ufor stream cipher is RC4 provides better security. Thus eachand every bit in the data is encrypted and is highly secured.3.2 Compression using DFT and Data EmbeddingCompression is a technique for reducing the file size making itsuitable for data transmission. There are two types ofcompression, namely lossless and lossy. Image in its 2-D formcan be compressed using various techniques. One primarymethod is to use Discrete Fourier transform (DFT). DFT givesthe better approximation of Fourier transform on discrete setof frequencies. The spatial values in such 2-dimension isrepresented as f(x,y). The transformed image is represented asF(u,v).ki,j,uHere i,j represents the row and column of the image and urepresents the face value of each bit in the pixel. In thisencryption, each bit of the pixel is XOR-ed with the keygenerated by the pseudorandom generator. As the encryptionkey is pseudo randomly generated and reproduced only at thereceiving end, which cannot be produced by any intruders andis difficult to break or analyze the image. A basic algorithm𝐹 𝑢, 𝑣 1𝑀. 𝑁𝑀 1 𝑁 1𝑢𝑥 𝑣𝑦𝑓(𝑥, 𝑦)𝑒 𝑗 2𝜋 ( 𝑀 𝑁 )𝑥 0 𝑦 0The inverse Fourier transform is thus given by,𝑓 𝑥, 𝑦 1𝑀. 𝑁𝑀 1 𝑁 1𝑢𝑥 𝑣𝑦𝐹(𝑢, 𝑣)𝑒 𝑗 2𝜋 ( 𝑀 𝑁 )𝑢 0 𝑣 0Volume: 03 Issue: 02 Feb-2014, Available @ http://www.ijret.org201

IJRET: International Journal of Research in Engineering and TechnologyeISSN: 2319-1163 pISSN: 2321-7308Case 1Encrypted imagecontainingembedded dataData ExtractionAdditional DataData Hiding KeyCase 2Encrypted imagecontainingembedded dataImage DecryptionDecrypted ImageEncryption KeyFig 3 Separable Receiver SectionThe transform equation suggests that an image of size MxNwith spatial values (x,y) is transformed into one with values(u,v) [4].Fast Fourier Transform (FFT) is the fastest method of DFT.FFT is a high speed and efficient technique. At thereconstruction stage, the image compressed using losslessmethod will have better image quality than the one that iscompressed using lossy method [5].3.3 Decryption and ExtractionIn the receiver side, the reverse process is carried out [3]. Hereas the data is received they can be separately extractedindependently using either the encryption key which is pseudorandomly generated to extract the decrypted image. Here thesame pseudo random generated must be provided by thereceiver in order to extract the cover image. Then the datahiding key is applied to the recovered image to extract thesecret data. The keys can be used alternatively to recover anyone of the data also. Hence offer better flexibility.(b) Stream Ciphered Image4. EXPERIMENTAL RESULTS(c) Secret Image(a) Original Cover ImageVolume: 03 Issue: 02 Feb-2014, Available @ http://www.ijret.org202

IJRET: International Journal of Research in Engineering and TechnologyeISSN: 2319-1163 pISSN: 2321-7308REFERENCES(d) Stego Image Containing Secret Image[1]. C. K. Chan and L. M. Cheng, “Hiding data in images bysimple LSB substitution,” Pattern Recognit., vol. 37, no. 3, pp.469–474, 2004.[2]. Manikandan R Asst. Professor, Uma M ,MahalakshmiPreethi S M, “Reversible Data Hiding forEncrypted Image”, Journal of Computer Applications ISSN:0974 – 1925,Volume-5, Issue EICA2012-1, February 10, 2012[3]. X. Zhang, “Reversible data hiding in encrypted image,”IEEE Signal Process. Lett., vol. 18, no. 4, pp. 255–258, Apr.2011.[4]. Inderjit Singh, Sunil Khullar, Dr. S.C. Laroiya, “DFTBased Image Enhancement and Steganography”, InternationalJournal of Computer Science and CommunicationEngineering, ISSN : 2319-7080, Vol2, Issue 1, Feb 2013.[5]. Mridul Kumar Mathur, Gunjan Mathur, “ ImageCompression using DFT through Fast Fourier TransformTechnique”, International Journal of Emerging Trends andTechnology in Computer Science”, ISSN : 2278-6856, Vol 1,Issue 2, July-Aug 2012.(e) Extracted Secret ImageFig. 4 Implemented OutputsThe figure above shows the results of the proposed system.Initially, an image taken is considered to be a cover image.This image is now given for DFT compression after which thesecret image is embedded into the cover. The obtained stegoimage is then given into a separable receiver for separableextraction of original image and the secret image.CONCLUSIONSIn this paper we briefly discuss about how to encrypt a data,compress the encrypted data and embed a data to thecompressed one. The later p