Slideshare uses cookies to improve functionality and performance, and to provide you with relevant advertising. If you continue browsing the site, you agree to the use of cookies on this website. See our User Agreement and Privacy Policy.

Slideshare uses cookies to improve functionality and performance, and to provide you with relevant advertising. If you continue browsing the site, you agree to the use of cookies on this website. See our Privacy Policy and User Agreement for details.

Like this document? Why not share!

660 views

Published on

No Downloads

Total views

660

On SlideShare

0

From Embeds

0

Number of Embeds

2

Shares

0

Downloads

26

Comments

0

Likes

1

No embeds

No notes for slide

- 1. B.Raja Rao, Dr.E.V.V.Krishna Rao, S.V.Rama Rao,M.Rama mohan rao / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 6, November- December 2012, pp.1598-1604 Finger Print Parameter Based Cryptographic Key Generation B.Raja Rao, Dr.E.V.V.Krishna Rao, S.V.Rama Rao,M.Rama mohan rao (Associate Professor, Department of ECE, URCE, Telaprolu,vijayawada. (Professor & H.O.D, Department of ECE, ALIET, vijayawada. (Associate Professor, Department of ECE, NRIIT,vijayawada. (Associate Professor & H.O.D, ECE, URCE, Telaprolu,vijayawada.ABSTRACT 3. Large size of strong keys results in longer delay in A method is proposed for generation of encryption/decryption.unique cryptographic key which is generated using 4. It is difficult to remember the keys, storing them infigure prints of the user, which are stable a data base may be insecure.throughout persons lifetime. The proposed 5. Moreover, maintaining and sharing lengthy,approach reduces the cost associated with lost random keys is the critical problem in thekeys, addresses non-repudiation issues and cryptography system.provides increased security of digital content. This A number of biometric characteristics areapproach has reduced the complicated sequence of being used in various applications as Universality,the operation to generate crypto keys as in the Uniqueness, Permanence, Measurability,traditional cryptography system . The key is Performance, Acceptability, and Circumvention [2].derived directly from the figure print data and isnot stored in the database, since it creates more 1.1 Biometric Based Cryptographycomplexity to crack or guess the cryptographic Biometrics and cryptography are twokeys. Biometrics, cryptography and data hiding potentially complementary security technologies.will provide good perspectives for information Biometrics gives a unique, measurable biologicalsecurity. We proposed an algorithm for deriving characteristic for automatically recognizing orthe key from biometric for ECC (Elliptic Curve verifying the identity of a human being. CryptographyCryptography) based applications which will is an important feature of computer and networkprovide high security with good performance in security. Using biometrics by means of cryptographyterms of computational and bandwidth is a new hot research topic. In this approach uniquerequirements.There are several biometric systems cryptographic key is derived directly from thein existence that deal with cryptography, but the biometric data of the user (i.e. fingerprint in thisproposed figure print parameter based approach).The encryption process begins with thecryptographic key introduces a novel method to acquisition of the required biometric samples.generate cryptographic key. This approach is Features and parameters are extracted from theseimplemented in MATLAB and can generate samples and used to derive a biometric key that can bevariable size cryptographic key, with minimum used to encrypt a plaintext message. The minutiaeamount of time complexity, which is aptly suited points are extracted from the fingerprint and that pointfor any real time cryptography. set is used for generating encryption key. Minutiae points are locations where a fingerprint ridge ends orKeywords - Cryptography, Biometrics, Minutiae bifurcates. There are several benefits of the proposedpoints, Morphological Operation, Histogram approach:Equalization, Crossing Number. 1. This Biometric based cryptosystems combine cryptography and biometrics to benefit from theI. INTRODUCTION strengths of both fields. With the widespread use of information 2. Provides increased security of digital content.exchange across the Internet, and the storage of 3. Biometrics brings in non repudiation. Allow onlysensitive data on open networks, cryptography is the legal user to utilize the content.becoming an increasingly important feature of data 4. The simplicity of use and the very limited risk ofsecurity. Many cryptographic algorithms are available losing, stealing or forging the user’s biologicalfor securing information E.g. RSA, DES, AES etc. identifier.Normally used cryptosystem [1] have a number ofassociated inconveniences and problems such as:- 1.2 Fingerprint1. Conventional Cryptography authenticates messages A fingerprint is the feature pattern of onebased on the key but not on the user. Hence unable to finger [3]. Each person has his own fingerprints withdifferentiate between the legitimate user & an the permanent uniqueness. However, shown byattacker. intensive research on finger print recognition,2. These keys can be guessed or cracked. fingerprints are not distinguished by their ridges and furrows, but by Minutia, which are some abnormal points on the ridges as shown in Fig. 1. Among the 1598 | P a g e
- 2. B.Raja Rao, Dr.E.V.V.Krishna Rao, S.V.Rama Rao,M.Rama mohan rao / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 6, November- December 2012, pp.1598-1604variety of minutia types reported in literature, two aremostly significant and in heavy usage: one is calledtermination, which is the immediate ending of a ridge;the other is called bifurcation, which is the point on lake hook bridge Double ifurcationthe ridge from which two branches derive. Valley isalso referred as Furrow, Termination is called asEnding, and Bifurcation is also called as Branch. The general shape of the finger print isgenerally used to pre-process the images .The firstscientific study of figure print was made by Galtonwho divided the figure print into 3 classes.ie Loop, Trifurcation opposed Ridge crossing opposedwhorl and arch. bifurcations ridge ending fig.2. Minutiae types with examples. 1.3 Unique Features Of Fingerprint : 1. Each fingerprint is unique to individuals i.e. no two fingers have identical ridge characteristics. 2. They are highly universal as majority of the a b c population have legible fingerprints. 3. They are very reliable as no two people have same fingerprint. Even identical twins having similar DNA, are believed to have different fingerprints. 4. Fingerprints are formed in the fetal stage and remain structurally unchanged throughout an individual’s lifetime. d e f 5. It is one of the most accurate forms of biometrics Fig.1.finger print types (a) arch (b) right loop (c) available. left 6. Fingerprint acquisition is non intrusive and hence is loop (d) double loop (e) whorl (f) tent arch a good option. The loop is by far the most common type offigure print .The human population has figure prints II . CRYPTOGRAPHIC KEY GENERATIONin the following percentages. FROM FINGER PRINT BIOMETRIC In our approach we have selected fingerprint Loop – 65 % Whorl -30 % Arch - as the biometrics feature for generating cryptographic05%. key. We have extracted minutiae points from the In this approach fingerprint is used as a fingerprint and that point set is used for generatingbiometric parameter for generation of encryption key. cryptographic key.Fingerprints have been used for over a century and arethe most widely used form of biometric identification. Finger Print image EnhancementThe fingerprint of an individual is unique and remainsunchanged over an individual’s lifetime. A fingerprintis formed from an impression of the pattern of ridgeson a finger. A ridge is valley is the region between Histogram Equilizationtwo adjacent ridges. The set of minutiae types arerestricted into only two types, ridge endings andbifurcations, Ridge endings are the points where the Binarizationridge curve terminates, and bifurcations are where aridge splits from a single path to two paths at a Y-junction. Figure 1 illustrates an example of a ridgeending and a bifurcation. In this example, the black Thinningpixels correspond to the ridges, and the white pixelscorrespond to the valleys. Minutiae Extractor (Xi , Yi ,θi )Minutiae examples : Key Generator (FPCONCAT)Bifurcation Ridge ending dot island fig.3.Block diagram of key generation system 1599 | P a g e
- 3. B.Raja Rao, Dr.E.V.V.Krishna Rao, S.V.Rama Rao,M.Rama mohan rao / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 6, November- December 2012, pp.1598-1604 binarize the fingerprint image. Such a named method The various steps required for generating comes from the mechanism of transforming a pixelcryptographic key from fingerprint biometric are value to 1 if the value is larger than the mean intensity 1. Fingerprint Image Enhancement value of the current block (16x16) to which the pixel 2. Histogram Equalization belongs 3. Fingerprint Image Binarization 2.4.Thinning 4. Thinning Thinning is the process of reducing the 5. Minutiae Extraction thickness of each line of patterns to just a single pixel 6. Key generation width The requirements of a good thinning algorithm with respect to a finger print are 2. 1.Finger print Image Enhancement a) The thinned fingerprint image obtained should be Fingerprint Image Enhancement is to make of single pixel width with no discontinuities.the image clearer for easy further operations. Since b) Each ridge should be thinned to its centre pixel.the fingerprint images acquired from sensors or c) Noise and singular pixels should be eliminated.other media are not assured with perfect quality, d) No further removal of pixels should be possibleenhancement methods, for increasing the contrast after completion of thinning process.between ridges and furrows and for connecting the The final image enhancement step typicallyfalse broken points of ridges due to insufficient performed prior to minutiae extraction is thinning [9].amount of ink, are very useful to keep a higher Thinning is a morphological operation thataccuracy to fingerprint recognition[5]. The quality of successively erodes away the foreground pixels untilthe ridge structures in a fingerprint image is an they are one pixel wide. A standard thinningimportant characteristic, as the ridges carry the algorithm is employed, which performs the thinninginformation of characteristic features required for operation using two sub iterations. This algorithm isminutiae extraction. Ideally, in a well-defined accessible in MATLAB via the `thin operation underfingerprint image, the ridges and valleys should the bwmorph function. Each sub iteration begins byalternate and flow in locally constant direction. This examining the neighborhoods of each pixel in theregularity facilitates the detection of ridges and binary image, and based on a particular set of pixel-consequently, allows minutiae to be precisely deletion criteria, it checks whether the pixel can beextracted from the thinned ridges. However, in deleted or not. These sub iterations continue until nopractice, a fingerprint image may not always be well more pixels can be deleted. The application of thedefined due to elements of noise that corrupt the thinning algorithm to a fingerprint image preservesclarity of the ridge structures. This corruption may the connectivity of the ridge structures while formingoccur due to variations in skin and impression a skeletonised version of the binary image. Thisconditions such as scars, humidity, dirt, and non- skeleton image is then used in the subsequentuniform contact with the fingerprint capture device. extraction of minutiae.Thus, image enhancement techniques are oftenemployed to reduce the noise and enhance the 2.5 Minutiae Extractiondefinition of ridges against valleys. The fingerprint recognition problem can be grouped into three sub-domains: fingerprint2.2. Histogram Equalization enrollment, verification and fingerprint Histogram equalization is to expand the pixel identification. In addition, different from the manualvalue distribution of an image so as to increase the approach for fingerprint recognition by experts, theperceptional information [6]. The original histogram fingerprint recognition here is referred as AFRSof a fingerprint image has the bimodal type, the (Automatic Fingerprint Recognition System), which ishistogram after the histogram equalization occupies program-based Verification is typically used forall the range from 0 to 255 and the visualization effect positive recognition, where the aim is to preventis enhanced. multiple people from using the same identity. Finger print verification is to verify the authenticity of one2.3. Fingerprint Image Binarization person by his fingerprint. There is one-to-one Most minutiae extraction algorithms operate comparison in this case. In the identification mode,on binary images where there are only two levels of the system recognizes an individual by searching theinterest: the black pixels that represent ridges, and the templates of all the users in the database for a match.white pixels that represent valleys. Binarization is the Therefore, the system conducts a one to-manyprocess that converts a grey level image into a binary comparison to establish an individual’s identity. Theimage. This improves the contrast between the ridges Following Techniques are used for Finger printand valleys in a fingerprint image, and consequently Recognitionfacilitates the extraction of minutiae. The outcome is abinary image containing two levels of information, the 2.5.1.Minutiae Extraction Technique Most of theforeground ridges and the background valleys. A finger-scan technologies are based on Minutiae.locally adaptive binarization method is performed to Minutia-based techniques represent the finger print by 1600 | P a g e
- 4. B.Raja Rao, Dr.E.V.V.Krishna Rao, S.V.Rama Rao,M.Rama mohan rao / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 6, November- December 2012, pp.1598-1604its local features, like terminations and bifurcations. Table I: Crossing Number PropertiesThis approach has been intensively studied, also is the S.no Property Crossing Numberbackbone of the current available fingerprint 1 Isolated point 0recognition products [4]. This work also concentrates 2 Ridge ending point 1on same approach. 3 Continuing ridge 2 point2.5.2. Pattern Matching or Ridge Feature Based 4 Bifurcation point 3Techniques 5 Crossing point 4 Feature extraction and template generation The Crossing Number (CN) method is usedare based on series of ridges as opposed to discrete to perform minutiae extraction. This method extractspoints which forms the basis of Pattern Matching the ridge endings and bifurcations from the skeletonTechniques. The advantage of Pattern Matching image by examining the local neighborhood of eachtechniques over Minutiae Extraction is that minutiae ridge pixel using a 3×3 window. The CN for a ridgepoints may be affected by wear and tear and the dis pixel P is given byadvantages are that these are sensitive to properplacement of finger and need large storage fortemplates.2.5.3. Correlation Based Technique Two figure prints are superimposed and P4 P3 P2 0 0 1 1 1 0correlation between corresponding pixels is computedfor different alignments. The most commonly employed method of P5 P P1 0 1 0 1 1 0minutiae extraction is the Crossing Number (CN)concept. This method involves the use of the P6 P7 P8 0 0 0 0 0 0skeleton image where the ridge flow pattern is eight-connected. The minutiae are extracted by scanning thelocal neighborhoods of each ridge pixel in the image a b cusing a 3×3 window. The CN value is then computed,which is defined as half the sum of the differences where Pi is the pixel value in the neighborhood of P.between pairs of adjacent pixels in the eight-neighborhoods. Using the properties of the CN as For a pixel P, its eight neighboring pixels are scannedshown in Table I below, the ridge pixel can then be in an anti- clockwise direction as followsclassified as a ridge ending, bifurcation or non-minutiae point. For example, a ridge pixel with a CN If the central is one-value and has only oneof one corresponds to a ridge ending, and a CN of one-value as neighbor, then it is an end pointthree corresponds to a bifurcation.. like in Figure b). A method of securing communication is If the central is one-value and has three one-proposed which overcomes several problems value as neighbor, then it is an bifurcationassociated with traditional cryptography. It provides a like in Fig c).practical and secure way to integrate the fingerprint After the CN for a ridge pixel has been computed, thebiometric into cryptographic applications. The crypto pixel can then be classified according to the propertykeys have been generated reliably from genuine of its CN value. As shown in Table I, a ridge pixelfingerprint samples, which is stable throughout with a CN of one corresponds to a ridge ending, and aperson’s lifetime[7]. This approach has reduced the CN of three corresponds to a bifurcation. For eachcomplicated sequence of the operation to generate extracted minutiae point, the following information iscrypto keys as in the traditional cryptography system recordedand hence requires minimum amount of time a) X and Y coordinates,complexity, which is aptly suited for any real time b) Orientation of the associated ridge segment, andcryptography. Provides increased security of digital c) Type of minutiae (ridge ending or bifurcation).content. Biometrics brings in non repudiation andallow only the legal user to utilize the content. Thereis very limited risk of losing, stealing or forging theuser’s biological identifier. 1601 | P a g e
- 5. B.Raja Rao, Dr.E.V.V.Krishna Rao, S.V.Rama Rao,M.Rama mohan rao / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 6, November- December 2012, pp.1598-1604Table II : Minutiae Points MB i = X Location (9bits) + X Location (9bits) Minutiae Position Minuti Minuti +angle(3 bits) ae type ae Step 3: Convert the above concatenated binary string Directi MBi to decimal to get the single co-ordinate value on M1i.X- Y– Crossi Theta Let K be a large enough integer so that theCoordinate Coordinate ng failure probability of 1 out of 2k will be satisfied Numbe when the plaintext m is attempted to imbed; in r practice k=30. Suppose message units are integers 033 238 3 270 <= m <= M, then finite filed should be selected to174 335 1 135 satisfy q>Mk. Therefore the integers are from 1 to Mk394 327 1 45 in the form mk+j, where 1<=j<=k [10]. Thus the503 537 3 225 given m, for each j = 1,2,…,k obtain an element x of698 112 1 90 Fp corresponding to mk+j by using the following856 381 3 270 equation.2.6. Cryptographic Key Generation Y2=x 3 +ax+b …….. (1) Elliptic curves are mathematical constructsthat have been studied by mathematicians since the Step 1: Let k = 30 Step 2: X i= MB i + j where 1 <=j<=kseventeenth century. The following section will Step 3. For each Xi, compute Yi using the followingprovide the general architecture of implementationand details about the algorithms for embedding equation.minutiae on elliptic curve and the biometric based key F(Xi) = Yi 2 = Xi 2 + aX i + b mod p ……(2)generation method. To encode plaintexts as points onsome elliptic curve E defined over a finite field Fp. Step 4: If f(X i) is non square, then increment j by 1Here plaintext is nothing but the minutiae co-ordinate and trywhich is extracted from the fingerprint. Minutiae isrepresented in three co-ordinate system as (x,y,θ). To again with new X i.map the minutiae on to the elliptic curve, this threeco-ordinate minutiae are converted into one co- Step 5: If j reaches K, then increment kordinate system and then this single co-ordinate valueis mapped on to elliptic curve.Therefore in the Step 6: Repeat the above steps for all minutiaeproposed algorithm, by using these co-ordinates, the Step 7: Mapped points Pmi € E(F q)private key is generated. To generate the key, all the Thus the single co-ordinate value is mapped on to theminutiae co-ordinates of the given image are to beextracted. elliptic curve.2.6.1. Algorithm III. Experimental Results First, add all x co-ordinates, y co-ordinates In real time application, it is applied in suchand θ co-ordinates separately and then take average a way that the enhanced finger print is given as inputvalue each co-ordinates separately. This resultant to the Minutiae Extractor. Output of Minutiaeaverage values of X and Y co-ordinates are in decimal Extractor is the list of Minutiae Co-coordinators,representation and the co-ordinate θ will be in radians. which is given as input to the Key Generator, whichNow this average values are converted into binary will generates the biometric based key.. More thanstring separately subject to the condition that the input 100 finger print images were tested using differentimage should be resized into 256x256 array during ECC parameter sets. The average computation timepreprocessing of the image for minutiae extraction. So taken for the key generation process is 0.032 ms. Thismaximum value for each co-ordinate require 9 bits for scheme is programmed in Matlab (Matlab7.6). Weeach X and Y co-ordinates to represent 256 and θ have tested the proposed system with diverserequire maximum of 3 bits. Finally, these three binary fingerprint images. The minutiae points are extractedstrings are converted into one co-ordinate value by from the fingerprint images using the approachconcatenation these three binary strings, which will discussed. A Unique cryptographic key is generatedgive the private key for the given finger printimage. from the biometric template of a user. The followingThe algorithm FPCONCAT for the above processes are the experimental results obtained for the proposedfollows as below: approachStep 1: Get the binary values XBi, YBi and θBi of Xi,Yi and θi for given i th minutiae.Step 2: Concatenate all the binary values in thefollowing order. 1602 | P a g e
- 6. B.Raja Rao, Dr.E.V.V.Krishna Rao, S.V.Rama Rao,M.Rama mohan rao / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 6, November- December 2012, pp.1598-1604 1100000000000011101000001100001000000000001 1011111011000001100101110010010111100000011 10010000010110000110000010000000001010100 fig. 8. Generated key. IV.CONCLUSION A method of securing communication is proposed which overcomes several problems associated with traditional cryptography. It provides a a b practical and secure way to integrate the fingerprint biometric into cryptographic applications. The crypto keys have been generated reliably from genuine fingerprint samples, which is stable throughout person’s lifetime. This approach has reduced the complicated sequence of the operation to generate crypto keys as in the traditional cryptography system and hence requires minimum amount of time complexity, which is aptly suited for any real time c cryptography[11]. Since ECC is adopted, this fig 4 a) Image of sample fingerprint b) Image after architecture may be suitable for mobile based histogram equalization c) Minutiae Points biometric applications. References [1] Umut Uludag, Sharath Pankanti, Salil Prabhakar, Anil K.Jain “Biometric Cryptosystems Issues and Challenges” Proceedings of the IEEE 2004. [2] A. Jain, R. Bolle, and S. Pankanti, “Biometrics Personal Identification in Networked Society”, Kluwer Academic Publishers New York, Boston, Dordrecht,. Original Image Enhanced image London, Moscow, pp. 1-64, 2002. fig. 5 Histogram Enhancement [3] P. Komarinski, P. T. Higgins, and K. M. Higgins, K. Fox Lisa,“Automated Fingerprint Identification Systems (AFIS)”, Elsevier cademic Press, pp. 1-118, 2005. [4] D. Maltoni, D. Maio, and A. Jain, S. Prabhakar, “4.3: Minutiae-based Methods’ (extract) from Handbook of Fingerprint Recognition”,Springer, New York, pp. 141- 144, 2003. Binarized image Enhanced gray image [5] N.Lalithamani, K.P.Soman “Irrevocable Cryptographic Key Generation from fig. 6. Finger print image after Binarization Cancelable Fingerprint Templates: An Enhanced and Effective Scheme”. European Journal of Scientific Research ISSN 1450- 216X Vol.31 No.3 (2009), pp.372-387 [6] K. Nallaperumall, A. L. Fred, and S. Padmapriya, “A Novel Technique for Fingerprint Feature Extraction Using Fixed Size Templates”, IEEE2005 Conference, pp. 371-374, 2005. [7] P. Komarinski, P. T. Higgins, and K. M. Higgins, K. Fox Lisa, “Automated Fingerprint fig. 7. Finger print image after Thinning Identification Systems (AFIS)”, ElsevierAcademic Press, pp. 1-118, 2005. The key generated from the sample finger [8] Y. Seto, “Development of personal print is authentication systems using fingerprint with smart cards and digital signature technologies,” the Seventh International 1603 | P a g e
- 7. B.Raja Rao, Dr.E.V.V.Krishna Rao, S.V.Rama Rao,M.Rama mohan rao / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 6, November- December 2012, pp.1598-1604 Conferenceon Control, Automation, Robotics nd Vision, Dec 2002 [9] L.C. Jain, U. Halici, I. Hayashi, S.B. Lee, and S. Tsutsui, “Intelligent biometric techniques in fingerprint and face recognition”, The CRC Press, 1999. [10] Koblitz.N A course in Number Theory and Cryptography, New York, Springer Verlog, Second Edition, 1994. [11] L.Lam, S.W.Lee and C.Y.Suen, “Thinning Methodologies – A comprehensive study”, IEEE Transaction on Pattern Analysis and Machine Intelligence, vol. 14,no 9,1992. 1604 | P a g e

No public clipboards found for this slide

×
### Save the most important slides with Clipping

Clipping is a handy way to collect and organize the most important slides from a presentation. You can keep your great finds in clipboards organized around topics.

Be the first to comment