Improved cancelable fingerprint templates using minutiae-based functional transform

Size: px
Start display at page:

Download "Improved cancelable fingerprint templates using minutiae-based functional transform"

Transcription

1 SECURITY AND COMMUNICATION NETWORKS Security Comm. Networks 2014; 7: Published online 17 May 2013 in Wiley Online Library (wileyonlinelibrary.com)..788 SPECIAL ISSUE PAPER using minutiae-based functional transform Daesung Moon 1, Jang-Hee Yoo 1 and Mun-Kyu Lee 2 * 1 ETRI, Korea 2 School of Computer and Information Engineering, Inha University, Korea ABSTRACT Since Ratha et al. introduced the functional transform for cancelable fingerprint templates, a few simulation attacks to this method have been proposed. The attacks are based on the fact that the transform depends only on the predefined parameters. That is, the attacker may fully simulate the transform and partially invert it if the parameters are available. Although an original template is transformed using different parameters for different systems, even the compromise of only one of these systems may reveal the original template, which may be a serious potential threat from a practical viewpoint. In this paper, we propose an improved functional transform, whose parameters are decided by the original template, as well as predefined user-specific keys. Because the information on the original template will not be available to the attacker even when a system is compromised, the proposed method significantly improves the security of the original template by preventing the attacker from reconstructing the transform. Copyright 2013 John Wiley & Sons, Ltd. KEYWORDS user authentication; cancelable biometrics; cancelable fingerprint template; functional transform *Correspondence Mun-Kyu Lee, School of Computer and Information Engineering, Inha University, Korea. mklee@inha.ac.kr 1. INTRODUCTION Cancelable biometrics [1,2] is a promising solution to security and privacy concerns about biometrics. The system does not store the original biometric data in this method, but it only stores the templates transformed by a noninvertible transform. Then, authentication is carried out on this transformed biometric data without a need to access the original data. Note that in this setting, a user s biometric data should remain safe, even if the system is compromised. The user just may cancel the old transformed biometric data, and enroll in the system with a new one, by another transform. Since Ratha et al. published their seminal paper on cancelable biometrics [1], various schemes for cancelable biometrics have been introduced. These include several reviewed in [3] and more recent ones, such as Teoh et al. s Biohashing [4], Ang et al. s key-based transform [5], Uludag et al. s fuzzy fingerprint vaults [6], and Lee et al. s local minutiae-based scheme [7]. The best known scheme for cancelable fingerprint templates is Ratha et al. s method [8], which transforms a fingerprint template into a completely different one using predefined parameters. However, it was shown that the original fingerprint template may be recovered if a transformed template, and the parameters used for this transformation are compromised [9,10]. The idea given in [8], that a single template may be transformed into various forms using different parameters, is novel and appropriate, but its implementation should be improved to fully accommodate the idea. The attacks presented in [9,10] use the fact that the knowledge of transformation parameters enables an attacker to partially invert the transform. The problem is that the transformation depends only on the predefined parameters that are revealed to the attacker when the system is compromised. Therefore, we modify the scheme so that a transformation should depend on some information not directly accessible by an attacker even if the system is compromised. An obvious example of this kind of information is the original template. Therefore, we design our transformation method so that both the minutiae positions in the original fingerprint template and the predefined parameters contribute to the shapes of transformation functions. Throughout this paper, a minutia means a fingerprint feature point composed of four elements, that is, x and y coordinates, angle, and type. Therefore, a minutia can be represented by m i =(x i,y i,θ i,t i ). A tricky point in this approach is that only a small Copyright 2013 John Wiley & Sons, Ltd. 1543

2 perturbation in the original template may cause a big change in the transformed one, because the minutiae positions play the two roles at the same time; they are the input to the transformation function, and they decide the function itself. To prevent this amplification of errors, we also propose various compensation techniques. The remainder of this paper is organized as follows. In Section 2, we review Ratha et al. s functional transform [8] and known attacks to it. Section 3 provides our improved functional transform as well as several refinement techniques. The performance of these techniques is evaluated in Section 4. Section 5 concludes the paper with suggestion for future work. 2. REVIEW OF RATHA ET AL. S FUNCTIONAL TRANSFORM Ratha et al. [8] introduced three methods to irreversibly transform a fingerprint template, that is, Cartesian, polar, and surface folding transforms. They mentioned that the third method is the most practical; thus, we concentrate on the surface folding transform, a.k.a. the functional transform, in this paper. Ratha et al. s functional transform is a one-way transform that moves minutiae positions according to the function defined by the predefined parameter set specific to each user. That is, the transform is described as a function over the feature domain that takes as input, the coordinates (x,y) of a possible minutia position and produces as output the amount and the direction of shift at this position. In [8], Gaussian mixtures generated by accumulating twodimensional Gaussian kernels were used for this purpose. The centers and shapes of these Gaussian kernels are parameterized by the unique key associated with each user. To be precise, a Gaussian mixture Fz! ðþfor a position vector z =[x,y] T is defined as follows. First, its magnitude is defined as Fz ðþ¼ X pi! i j2pλij e 1 2 ðz m i Þ T Λ 1 i ðz m Þ i ; where the weight p i, covariance Λ i, and center m i of each Gaussian kernel are parameters given by the key. Then, the phase is n o defined as Φ F ðþ z ¼ 1arg r! 2 Fz ðþ þ Φ rand ; where Φ rand is the phase offset specified by the key. In [8], the authors used two separate functions! F and! G to avoid any form of correlation between the direction and the amount of a shift. Thus, a transformation (x,y,θ)! (X,Y,Y) isdefined by X ¼ x þ K! G ðx; yþ þ KcosðΦ F ðx; yþþ; (1) Y ¼ y þ K! G Y ¼ ðx; y Þ þ KsinðΦ F ðx; yþþ; (2) ðθ þ Φ G ðx; yþ þφ rand Þ mod 2p; (3) where K is a predefined constant. Figure 1 shows an example of a Gaussian mixture produced by placing 24 Gaussian distributions in the image space, where each distribution has a standard deviation of 50 pixels as in [8]. The magnitudes of! F and! G are scaled so that the highest and lowest peaks have values of 1 and 0, respectively. A fingerprint template may be transformed into a completely different one using two Gaussian mixtures, as shown in Figure 2. The rationale for this design is that the function should have a folding effect, that is, there are transformed regions that originate from multiple locations in the original space to satisfy the noninvertibility requirement. It was claimed in [8] that thanks to this noninvertibility property, an attacker cannot recover the original template, even if he or she obtains the transformed template, as well as the parameters, that is, the user-specific keys. However, given K,!! F, and G, an attacker can simulate (1) and (2) to build a Figure 1. Example Gaussian mixture! F generated accumulating 24 two-dimensional Gaussian distributions. magnitude j! Fz ð Þj and phase Φ F ðzþ Security Comm. Networks 2014; 7: John Wiley & Sons, Ltd.

3 Original template Transformed template Figure 2. Transformation of a fingerprint template by surface folding transform based on two Gaussian mixtures. original template and transformed template. dictionary consisting of 4-tuples (x,y,x,y) that has elements. It is possible to find the original minutiae points (x,y,θ) by matching the transformed minutiae points with the dictionary, because the transformed template, that is, the set of 3-tuples (X,Y,Y), is also given to the attacker. Some points are not uniquely recovered because of the many-to-one property of the transform. However, Quan et al. [9] pointed out that most of the regions do not have any overlap with other regions in the transformed template, while they have distortion. We can verify this fact from Figure 5(c) in [8]. According to the experiments in [9] that were performed on FVC2002 DB1 [9], about 90% of the minutiae positions in the transformed template have only one preimage. Moreover, if the attacker may obtain two or more pairs of (key, transformed template) originating from the same original template by attacking multiple databases, then almost 100% of the minutiae positions are uniquely determined [10]. 3. CANCELABLE FINGERPRINT TEMPLATES BASED ON MODIFIED FUNCTIONAL TRANSFORM 3.1. Basic method We observe that the vulnerability of [8] stems from the fact that the transform depends only on the predefined parameters. We improve on Ratha et al. s scheme so that the minutiae points in the original template also contribute to the distribution of Gaussian kernels, thus, the shapes of transformation functions. As a result, the attacker cannot simulate the transformation even if a system is compromised, because the information on the original minutiae points will not be available to the attacker in this case. Our method also requires a key specific to each user as in the original method. However, this key does not directly determine the locations of Gaussian kernels, but it only provides the information to extract the coordinates of Gaussian kernels from the original template. To be precise, the key defines rectangular regions, that is, the positions and dimensions of rectangles, as shown in Figure 3. We use eight rectangles in this example and throughout this paper, although this may be customized for each application. Then, each rectangular region is expanded to fit into the region, and the original minutiae inside the rectangle are moved accordingly, as shown in Figure 4. For example, the marked minutia in the lower left corner in Figure 4 moves to the marked position in Figure 4. Repeating this procedure for the other three rectangles, we finally obtain four expanded images. Then, by overlapping the four images, we obtain a final image with more points, such as Figure 4(c). Each point in this figure serves as a center for a Gaussian distribution with the same magnitude, 1, and the same standard deviation of 50 pixels. In the original transformation [8], each Gaussian distribution has a predefined peak value of either + 1 or 1, a distinct weight p i and covariance Λ i. In our method, we give a peak value according to the type of each minutiae; + 1 for ridge endings and 1 for ridge bifurcations. However, we cannot specify the weight and the covariance for each Gaussian distribution because the minutiae are changed in every trial for fingerprint recognition and any specific Gaussian kernel cannot be named. Therefore, we give the same shape and weight to all distributions. According to our experiments, this change does not affect the overall performance of fingerprint verification. After we generate as many Gaussian distributions as the center positions shown in Figure 4(c), we accumulate these distributions to obtain! F. We repeat the same task for! G. Then, a functional transform is defined by the formulas (1), (2), and (3). For this to be possible, a user-specific key should contain the positions and sizes of eight rectangles, as well as the phase offset Φ rand. We see that it is impossible for an attacker to build a dictionary of (x,y,x,y), because the exact formulas (1), (2), and (3) for transformation cannot be obtained without the knowledge of coordinates of original minutiae. Therefore, the original template remains safe, even if a transformed template and the specifications of rectangles are revealed to the attacker from a compromised database. Security Comm. Networks 2014; 7: John Wiley & Sons, Ltd. 1545

4 Four rectangles to extract Guassian kernels for Four rectangles to extract Guassian kernels for Figure 3. Examples of rectangular regions to extract! F and! G, overlapped with a fingerprint template. four rectangles to extract Gaussian kernels for! F and four rectangles to extract Gaussian kernels for! G Parameter selection We need to be cautious in the selection of the rectangles, because they play a dominant role in our scheme. In fact, Figure 3 contains both good and bad examples of rectangles. We see that the narrow rectangle in the upper part of Figure 3 contains no minutia in the original template; this means that this rectangle does not contribute to the transformation. The same problems exist in another narrow rectangle in Figure 3 and a small rectangle in the upper part of Figure 3. Examining these bad rectangles, we see that the size of a rectangle matters. That is, a small rectangle tends not to contain a sufficient number of minutiae. A straightforward solution to both of these problems is to use only big rectangles. However, using only big rectangles poses another problem. For example, consider an extreme case where all of the four rectangles almost span the entire region. Then, the overlapped image will be almost the same as the original template, which significantly limits the variation in the transform. Therefore, our final choice is to mix large and small rectangles. According to our extensive experiments, the optimal choice is to restrict the length of each side of a rectangle to between 128 and 134 pixels Alignment It is difficult to match two transformed templates, if their original minutiae positions have not been measured with regard to the same coordinate system, because functional transforms do not preserve geometric relations between minutiae. Therefore, it is preferable to pre-align the minutiae positions before transformation. Ratha et al. [8] address this issue using advanced techniques to determine core and delta points [12,13], as well as their own improvements on these techniques [14]. However, even these approaches cannot provide the exact alignment but only guarantee an approximate alignment. According to ([15], p. 236), core point(s) are difficult to reliably extract in a poor quality fingerprint. Moreover, due to variations in placement of the finger on the sensor, the core point may not always be imaged. Note that even a small error in alignment may be amplified to a big change in a transformed template. Especially in our modification, we should be more cautious, because the change in an original minutiae position will be accumulated in an erroneous function, as well as in an erroneous input. Therefore, we apply a more fault-tolerant technique, geometric hashing [16], that stores the minutiae information redundantly. To be precise, our registration phase should be as follows: Obtain a user s fingerprint template. Let T be this template, and let N be the number of feature points in T. Let P 1...P N be the feature points. Make N copies of T. Let T 1...T N be these copies. For i =1 to N, shift and rotate T i so that P i may be located at the center of the image, and its orientation may be in parallel with the X-axis. Apply the functional transform separately to each of the modified templates. Let T T 0 N be the transformed templates. Store T T 0 N as well as the parameters for the transform. On the other hand, the matching phase is as follows: Obtain a user s fingerprint template. Let U be this template, and let M be the number of feature points in U. Note that M may not be the same as N even if the user is the legitimate owner of T. Let Q 1...Q M be the feature points. Make M copies of U, and let U 1...U M be these copies. For j =1 to M, shift and rotate U j so that Q j may be located at the center of the image, and its orientation may be in parallel with the X-axis. Apply the functional transform separately to each of the modified templates by using the parameter 1546 Security Comm. Networks 2014; 7: John Wiley & Sons, Ltd.

5 (c) (d) Figure 4. Extracting centers of Gaussian kernels and generating a transformed template. rectangle expanded to region, Feature points moved according to expansion. Points outside the rectangle are discarded, (c) four expanded regions overlapped to decide the centers of Gaussian kernels used for! F, and (d) repeating the same task for! G and applying the formulas (1), (2), and (3), we obtain the transformed template. associated to the claimed user. Let U U 0 M be the transformed templates. Try to match T 0 i and U 0 j for i =1...N and j =1...M. Find a pair (T 0 i,u 0 j) with the maximum matching score. If this score is above a predefined threshold, accept the user. We remark that in an environment where resources are highly constrained, we may select random subsets of {T 0 i} and {U 0 j} of reasonable sizes instead of using all combinations to speed up the matching process Quantization of minutiae positions In our method, the positions of minutiae decide the centers of Gaussian kernels. Simultaneously, they are input to the transform, which means that a small perturbation in the original template may be amplified through a transform. Figure 5 shows an extreme case of error amplification. This example is one of the pre-aligned templates according to the geometric hashing method explained in the previous section, where A is the reference point. Assume that a small error occurs when an image recognition procedure locates the minutiae B and C, and they are slightly misaligned to D and E, respectively, as in Figure 5. The rectangle is expanded to fit into the entire image, and the resulting image will contain A 0 and D 0 originating from A and D, respectively. (Throughout this paper, X 0 denotes a point in an expanded image that originates from X in the original image.) Then, one of the Gaussian distributions will be centered at D 0, which should be B 0 if correctly aligned. Figure 5 shows an overlapped view of the expanded image and the Security Comm. Networks 2014; 7: John Wiley & Sons, Ltd. 1547

6 Figure 5. Error amplification due to perturbation. minutiae B and C slightly misaligned to D and E, respectively and point E is farther from the center of a Gaussian distribution, D, than it should originally be. original template, where the points from the expanded image, that is, A 0 and D 0 are centers of Gaussian distributions and the points from the original template, that is, A, D, and E, are input to the functional transform. We see that E is quite far from D 0, whereas its correct location, C, should be very close to the correct center, B 0. Thus, the transformed template may be quite a different one from what it should be. We can partially solve this problem by quantizing minutiae positions before extracting Gaussian distributions from them. In our modification, we do not directly use the overlapped image, such as Figure 4(c), to generate Gaussian surfaces. Instead, we partition the overlapped image into square cells of the same sizes, where the size of each cell is a predefined parameter. This parameter may be the same for all users of a system, but it may be distinct to each user. Then for each cell, we check if there is any minutia position in it, and mark all the cells that have minutiae positions. Next, we delete all the points and locate a single point at the center of each marked cell. Even if there are multiple points in a cell, only a single representative point will be given to a single cell. If a point is on the edges of two neighboring cells, it should contribute to only one of them according to a predefined rule, for example, a right and upper cell may have the priority. Thus, the centers of Gaussian kernels are located only on a grid. Figure 6 illustrates the effect of quantization. We see that in this example, even if B 0 and C 0 are inappropriately placed, that is, as D 0 and E 0, respectively, the quantized points will remain correct, as shown in Figure 6. We remark that we also should modify the peak value of each Gaussian kernel to apply this technique. Assume that there are two points in a cell and they are a ridge ending and a ridge bifurcation, whose peak values are + 1 and 1, respectively. Thus, the peak value of the representative point of this cell is not defined. Therefore, we decide to use only one peak value, that is, + 1, for all Gaussian kernels. Partition template into same sized square cells. Points in each cell moved to representative point at the cell center Figure 6. Quantization of minutiae positions. partition template into same sized square cells and points in each cell moved to representative point at the cell center Security Comm. Networks 2014; 7: John Wiley & Sons, Ltd.

7 3.5. Duplicated quantization Quantization of minutiae positions mostly solves the problem of error amplification. However, this approach may cause a side effect in some rare cases. Figure 7 illustrates this situation. Assume that the point B 0 is misplaced into D 0.AsD 0 is in the neighboring cell, the error, that is, the distance between B 0 and D 0, will be increased after quantization to the distance between the centers of two neighboring cells. The same problem occurs for C 0 that is misplaced into E 0. We quantize each point twice to compensate this side effect. However, for the second quantization, the grids are relatively shifted from those for the first quantization. Figure 7 explains this idea. In this figure, the dashed lines generate the cells for the second quantization, whereas the gray solid lines are for the original quantization. Then, we merge the results of two quantizations, which doubles the number of Gaussian kernels. We see that even if B 0 and D 0 belong to two different cells in the original quantization, they belong to the same cell in the second quantization. Figure 7(c) and (d) shows the overall effects of this duplicated quantization. Figure 7(c) is the result of a correct quantization when we apply the duplicated quantization, where B 0 1, C 0 1 are from the original quantization and B 0 2, C 0 2 are from the second quantization. If B 0 and C 0 are incorrectly located into D 0 and E 0,asin Figure 7, the result will be like that in Figure 7(d). Note that B 0 2 = D 0 2 and C 0 2 = E 0 2. Now, we explain why the duplicated quantization approach may alleviate the side effect of quantization. First, we recall that all the points in Figure 7(c) and (d) will serve as the centers of two-dimensional Gaussian distributions. Then, our approach may be summarized as adding a correct Gaussian distribution, for example, centered at B 0 2 = D 0 2, near an incorrect distribution, for example, centered at D 0 1, to reduce the influence of incorrect (c) (d) Figure 7. Effects of duplicated quantization. quantization amplifies errors between B and D and between C and E, each point is quantized one more time to the cell centers generated by dashed lines, (c) quantized points after merging the results of two quantizations from B and C, and (d) quantized points after merging the results of two quantizations from D and E. Security Comm. Networks 2014; 7: John Wiley & Sons, Ltd. 1549

8 distribution. We give an example with one-dimensional Gaussian distributions in Figure 8 to help understand the rationale for this approach. Although this setting differs from that of two-dimensional distributions, the overall tendencies are sufficiently similar. Now, assume that there is only one minutia position x = 1 in our one-dimensional template. Then, G B (x) in Figure 8 represents the correct Gaussian distribution centered at this point. It is scaled so that its maximum value is 1. Assume that an error occurs and it is shifted to x = 1, producing G D (x). Then, the amount of error in functional transformation at a specific position x resulting from this shift can be represented as E 1 (x) =G B (x) G D (x). Now, assume that there are originally two Gaussian distributions G B (x) andg new (x) instead of just one, and G new (x) remains correct when G B (x) is shifted to G D (x). The correct functional transform is expressed as G B (x) +G new (x), because a functional transform is defined by overlapping all the Gaussian distributions. However, a scaling factor c should be multiplied to this function, so that it may have 1 as its maximum value. In this example, c 1/1.76. Figure 8 plots the resulting function, as well as its erroneous version c(g D (x) +G new (x)). We see that each of the merged functions c(g B (x) +G new (x)) and c(g D (x) +G new (x)) seems like an individual Gaussian distribution, and the distance between their peaks is much smaller than that between G B (x) andg D (x). Intuitively, this means the amount of erroneous shift of G B (x) has been reducedbyaddingg new (x). To be precise, we see that the new error function E 2 (x) = c(g B (x) +G new (x)) c(g D (x)+ G new (x)) = c(g B (x) G D (x)) should have a smaller value than the original error function E 1 (x) for all x, asc < 1. Returning to the two-dimensional setting in Figure 7, we see that although the Gaussian surface accumulating the Gaussian kernels centered at D 0 1, D 0 2, E 0 1, E 0 2 still differs from that of Figure 7(c), it significantly reduces the negative effect of D 0 1 and E 0 1. Finally, we remark that this technique can also be applied to alleviate the error amplification problem of the Cartesian and polar transformations in [8]. 4. EXPERIMENTAL RESULTS In this section, we evaluate the performance of our method using FVC2002 DB1 set A [11]. Our experiments were carried out according to the evaluation procedure in the FVC2002 competition. That is, each transformed template is matched against the remaining seven transformed templates of the same finger to compute the false non-match rate, and the first transformed template of each finger is matched against the first transformed templates of the remaining fingers to compute the false match rate. Thus, the numbers of genuine tests and false acceptance tests are 8 7/2 100 = 2800 and /2 = 4950, respectively. We used a Microsoft Windows based machine with a 2.66 GHz CPU and 3 GB RAM, and computed the equal error rate (EER) as a performance indicator. Table I compares the performance of our modified functional transform and various refined versions. The first row shows the EER of the geometric hashing without any functional transform, whereas the second row shows the value for our modified functional transform. As explained in Section 2, we restricted the length of each side of a rectangle to between 128 and 134 pixels. The third and the fourth rows show the EERs for the quantization method (Section 3.4) and the duplicated quantization method (Section 3.5), respectively. We used pixel cells for both quantization methods, because its performance is slightly better than other choices, such as and according to our experiments. We see from Table I that quantization notably improves the performance of our functional transform, and duplication also has a positive effect, although it is minor. In the quantization method, the total execution times of the false non-match rate and false match rate tests are approximately 1.8 and 1.9 s, respectively. Table I. EER comparison for functional transformation methods. Method EER (%) Figure 8. Effect of duplicated quantization in one-dimensional setting. We add a new distribution G new to reduce the error between G B and G D. Error significantly reduced from E 1 to E 2. No transform (geometric hashing) 4.5 Modified functional transform 12.4 Quantization (cell size: 16 pixels) 7.4 Duplicated quantization 7.2 EER, equal error rate Security Comm. Networks 2014; 7: John Wiley & Sons, Ltd.

9 5. CONCLUSION AND FUTURE WORKS In this paper, we presented a new functional transformation method to enhance the security of Ratha et al. s method [8]. In our method, the transformation function depends on the original fingerprint template, as well as predefined parameters, to prevent an attacker from building an attack dictionary, even if the system is compromised. On the basis of our experimental results, we verified that the proposed approach can perform fingerprint verification more securely without a significant degradation of verification accuracy. However, we remark that it would be still necessary to enhance the verification accuracy of our method for its broader acceptance, because our method does not achieve the same performance as biometric verification methods with no transformation. We found out that, in most cases, complete deletion of some part of the fingerprint image may have a negative effect on verification accuracy. A possible solution would be to use a convex hull of two transformed templates, instead of templates themselves for matching. However, a precise analysis should be carried out with this approach, because the convex hull of two transformed templates does not exactly reflect the convex hull of original templates. We leave this issue for our future work. ACKNOWLEDGEMENTS M.-K. Lee was supported by the IT R&D program of MKE/KEIT [ , Intuitive, convenient, and secure HCI-based usable security technologies for mobile authentication and security enhancement in mobile computing environments] and by the MKE/ITRC (Information Technology Research Center) support program (NIPA H ) supervised by the NIPA(National IT Industry Promotion Agency). REFERENCES 1. Ratha NK, Connell JH, Bolle RM. Enhancing security and privacy in biometrics-based authentication system. IBM Systems Journal 2001; 40(3): Lee Y, Lee Y, Chung Y, Moon K. Secure face authentication framework in open networks. ETRI Journal 2010; 32(6): Uludag U, Pankanti S, Prabhakar S, Jain AK. Biometric cryptosystems: issues and challenges. Proceedings of the IEEE 2004; 92(6): Teoh ABJ, Ngo DCL, Goh A. BioHashing: Two factor authentication featuring fingerprint data and tokenised random number. Pattern Recognition 2004; 37(11): Ang R, Safavi-Naini R, McAven L. Cancelable keybased fingerprint templates. In ACISP 2005, Vol of LNCS. Springer: New York, 2005; Uludag U, Pankanti S, Jain AK. Fuzzy vault for fingerprints. In AVBPA 2005, Vol 3546 of LNCS. Springer New York, 2005; Lee C, Choi J-Y, Toh K-A, Lee S, Kim J. Alignmentfree cancelable fingerprint templates based on local minutiae information. IEEE Transactions on Systems, Man, and Cybernetics - Part B: Cybernetics 2007; 37(4): Ratha NK, Chikkerur S, Connell JH, Bolle RM. Generating cancelable fingerprint templates. IEEE Transactions on Pattern Analysis and Machine Intelligence 2007; 29(4): Quan F, Fei S, Anni C, Feifei Z. Cracking cancelable fingerprint template of ratha. In 2008 International Symposium on Computer Science and Computational Technology, pages IEEE Computer Society, Shin SW, Lee M-K, Moon D, Moon K. Dictionary attack on functional transform-based cancelable fingerprint templates. ETRI Journal 2009; 31(5): Maio D, Maltoni D, Wayman JL, Jain AK. FVC2002: second fingerprint verification competition. In Int. Conf. Pattern Recognition (ICPR 2002), pages IEEE Computer Society, Nilsson K, Bigun J. Localization of corresponding points in fingerprints by complex filtering. Pattern Recognition Letters 2003; 24(13): Nilsson K. Symmetry filters applied to fingerprints, PhD thesis, Chalmers Univ. of Technology, Sweden. 14. Chikkerur S, Ratha NK. Impact of singular point detection on fingerprint matching performance. In Fourth IEEE Workshop on Automatic Identification Advanced Technologies, pages IEEE Computer Society, Ratha N, Bolle R. Automatic Fingerprint Recognition Systems. Springer: New York, Chung Y, Moon D, Lee S, Jung S, Kim T, Ahn D. Automatic alignment of fingerprint features for fuzzy fingerprint vault. In Information Security and Cryptology, First SKLOIS Conf., Vol of LNCS. Springer: New York, 2005; Security Comm. Networks 2014; 7: John Wiley & Sons, Ltd. 1551

Mahmood Fathy Computer Engineering Department Iran University of science and technology Tehran, Iran

Mahmood Fathy Computer Engineering Department Iran University of science and technology Tehran, Iran 1 Alignment-Free Fingerprint Cryptosystem Based On Multiple Fuzzy Vault and Minutia Local Structures Ali Akbar Nasiri Computer Engineering Department Iran University of science and technology Tehran, Iran

More information

Combined Fingerprint Minutiae Template Generation

Combined Fingerprint Minutiae Template Generation Combined Fingerprint Minutiae Template Generation Guruprakash.V 1, Arthur Vasanth.J 2 PG Scholar, Department of EEE, Kongu Engineering College, Perundurai-52 1 Assistant Professor (SRG), Department of

More information

Implementation of the USB Token System for Fingerprint Verification

Implementation of the USB Token System for Fingerprint Verification Implementation of the USB Token System for Fingerprint Verification Daesung Moon, Youn Hee Gil, Sung Bum Pan, and Yongwha Chung Biometrics Technology Research Team, ETRI, Daejeon, Korea {daesung, yhgil,

More information

Local Correlation-based Fingerprint Matching

Local Correlation-based Fingerprint Matching Local Correlation-based Fingerprint Matching Karthik Nandakumar Department of Computer Science and Engineering Michigan State University, MI 48824, U.S.A. nandakum@cse.msu.edu Anil K. Jain Department of

More information

Using Support Vector Machines to Eliminate False Minutiae Matches during Fingerprint Verification

Using Support Vector Machines to Eliminate False Minutiae Matches during Fingerprint Verification Using Support Vector Machines to Eliminate False Minutiae Matches during Fingerprint Verification Abstract Praveer Mansukhani, Sergey Tulyakov, Venu Govindaraju Center for Unified Biometrics and Sensors

More information

Finger Print Enhancement Using Minutiae Based Algorithm

Finger Print Enhancement Using Minutiae Based Algorithm Available Online at www.ijcsmc.com International Journal of Computer Science and Mobile Computing A Monthly Journal of Computer Science and Information Technology IJCSMC, Vol. 3, Issue. 8, August 2014,

More information

Reducing FMR of Fingerprint Verification by Using the Partial Band of Similarity

Reducing FMR of Fingerprint Verification by Using the Partial Band of Similarity Reducing FMR of Fingerprint Verification by Using the Partial Band of Similarity Seung-Hoon Chae 1,Chang-Ho Seo 2, Yongwha Chung 3, and Sung Bum Pan 4,* 1 Dept. of Information and Communication Engineering,

More information

A Geometric Transformation to Protect Minutiae-Based Fingerprint Templates

A Geometric Transformation to Protect Minutiae-Based Fingerprint Templates A Geometric Transformation to Protect Minutiae-Based Fingerprint Templates Yagiz Sutcu a, Husrev T. Sencar b and Nasir Memon b a Polytechnic University, Electrical & Computer Engineering Dept., Brooklyn,

More information

Ujma A. Mulla 1 1 PG Student of Electronics Department of, B.I.G.C.E., Solapur, Maharashtra, India. IJRASET: All Rights are Reserved

Ujma A. Mulla 1 1 PG Student of Electronics Department of, B.I.G.C.E., Solapur, Maharashtra, India. IJRASET: All Rights are Reserved Generate new identity from fingerprints for privacy protection Ujma A. Mulla 1 1 PG Student of Electronics Department of, B.I.G.C.E., Solapur, Maharashtra, India Abstract : We propose here a novel system

More information

Online and Offline Fingerprint Template Update Using Minutiae: An Experimental Comparison

Online and Offline Fingerprint Template Update Using Minutiae: An Experimental Comparison Online and Offline Fingerprint Template Update Using Minutiae: An Experimental Comparison Biagio Freni, Gian Luca Marcialis, and Fabio Roli University of Cagliari Department of Electrical and Electronic

More information

Fingerprint Mosaicking by Rolling with Sliding

Fingerprint Mosaicking by Rolling with Sliding Fingerprint Mosaicking by Rolling with Sliding Kyoungtaek Choi, Hunjae Park, Hee-seung Choi and Jaihie Kim Department of Electrical and Electronic Engineering,Yonsei University Biometrics Engineering Research

More information

Fingerprint Matching using Gabor Filters

Fingerprint Matching using Gabor Filters Fingerprint Matching using Gabor Filters Muhammad Umer Munir and Dr. Muhammad Younas Javed College of Electrical and Mechanical Engineering, National University of Sciences and Technology Rawalpindi, Pakistan.

More information

Logical Templates for Feature Extraction in Fingerprint Images

Logical Templates for Feature Extraction in Fingerprint Images Logical Templates for Feature Extraction in Fingerprint Images Bir Bhanu, Michael Boshra and Xuejun Tan Center for Research in Intelligent Systems University of Califomia, Riverside, CA 9252 1, USA Email:

More information

A Framework for Efficient Fingerprint Identification using a Minutiae Tree

A Framework for Efficient Fingerprint Identification using a Minutiae Tree A Framework for Efficient Fingerprint Identification using a Minutiae Tree Praveer Mansukhani February 22, 2008 Problem Statement Developing a real-time scalable minutiae-based indexing system using a

More information

Reference Point Detection for Arch Type Fingerprints

Reference Point Detection for Arch Type Fingerprints Reference Point Detection for Arch Type Fingerprints H.K. Lam 1, Z. Hou 1, W.Y. Yau 1, T.P. Chen 1, J. Li 2, and K.Y. Sim 2 1 Computer Vision and Image Understanding Department Institute for Infocomm Research,

More information

Development of an Automated Fingerprint Verification System

Development of an Automated Fingerprint Verification System Development of an Automated Development of an Automated Fingerprint Verification System Fingerprint Verification System Martin Saveski 18 May 2010 Introduction Biometrics the use of distinctive anatomical

More information

SECURING information and ensuring the privacy of personal

SECURING information and ensuring the privacy of personal IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. 29, NO. 4, APRIL 2007 561 Generating Cancelable Fingerprint Templates Nalini K. Ratha, Fellow, IEEE, Sharat Chikkerur, Student Member,

More information

FINGERPRINT VERIFICATION BASED ON IMAGE PROCESSING SEGMENTATION USING AN ONION ALGORITHM OF COMPUTATIONAL GEOMETRY

FINGERPRINT VERIFICATION BASED ON IMAGE PROCESSING SEGMENTATION USING AN ONION ALGORITHM OF COMPUTATIONAL GEOMETRY FINGERPRINT VERIFICATION BASED ON IMAGE PROCESSING SEGMENTATION USING AN ONION ALGORITHM OF COMPUTATIONAL GEOMETRY M. POULOS Dept. of Informatics University of Piraeus, P.O. BOX 96, 49100 Corfu, Greece

More information

Indexing Fingerprints using Minutiae Quadruplets

Indexing Fingerprints using Minutiae Quadruplets Indexing Fingerprints using Minutiae Quadruplets Ogechukwu Iloanusi University of Nigeria, Nsukka oniloanusi@gmail.com Aglika Gyaourova and Arun Ross West Virginia University http://www.csee.wvu.edu/~ross

More information

Local Feature Extraction in Fingerprints by Complex Filtering

Local Feature Extraction in Fingerprints by Complex Filtering Local Feature Extraction in Fingerprints by Complex Filtering H. Fronthaler, K. Kollreider, and J. Bigun Halmstad University, SE-30118, Sweden {hartwig.fronthaler, klaus.kollreider, josef.bigun}@ide.hh.se

More information

PIN-based cancelable biometrics

PIN-based cancelable biometrics PIN-based cancelable biometrics Patrick Lacharme, Aude Plateaux To cite this version: Patrick Lacharme, Aude Plateaux. PIN-based cancelable biometrics. International Journal of Automated Identification

More information

Exploring Similarity Measures for Biometric Databases

Exploring Similarity Measures for Biometric Databases Exploring Similarity Measures for Biometric Databases Praveer Mansukhani, Venu Govindaraju Center for Unified Biometrics and Sensors (CUBS) University at Buffalo {pdm5, govind}@buffalo.edu Abstract. Currently

More information

Outline. Incorporating Biometric Quality In Multi-Biometrics FUSION. Results. Motivation. Image Quality: The FVC Experience

Outline. Incorporating Biometric Quality In Multi-Biometrics FUSION. Results. Motivation. Image Quality: The FVC Experience Incorporating Biometric Quality In Multi-Biometrics FUSION QUALITY Julian Fierrez-Aguilar, Javier Ortega-Garcia Biometrics Research Lab. - ATVS Universidad Autónoma de Madrid, SPAIN Loris Nanni, Raffaele

More information

Cancelable Key-Based Fingerprint Templates

Cancelable Key-Based Fingerprint Templates Cancelable Key-Based Fingerprint Templates Russell Ang, Rei Safavi Naini, and Luke McAven School of Information Technology and Computer Science, University of Wollongong, Northfields Avenue, NSW 2522,

More information

AN AVERGE BASED ORIENTATION FIELD ESTIMATION METHOD FOR LATENT FINGER PRINT MATCHING.

AN AVERGE BASED ORIENTATION FIELD ESTIMATION METHOD FOR LATENT FINGER PRINT MATCHING. AN AVERGE BASED ORIENTATION FIELD ESTIMATION METHOD FOR LATENT FINGER PRINT MATCHING. B.RAJA RAO 1, Dr.E.V.KRISHNA RAO 2 1 Associate Professor in E.C.E Dept,KITS,DIVILI, Research Scholar in S.C.S.V.M.V

More information

Generation of Combined Minutiae Template for Enrollment and Fingerprint Authentication

Generation of Combined Minutiae Template for Enrollment and Fingerprint Authentication Generation of Combined Minutiae Template for Enrollment and Fingerprint Authentication Dr.G.S.ANANDHA MALA Professor & Head Dept. of CSE, St. Joseph s College of Engineering, Chennai-6001119. gs.anandhamala@gmail.com

More information

An Efficient Secure Biometric System with Non-Invertible Gabor Transform

An Efficient Secure Biometric System with Non-Invertible Gabor Transform www.ijcsi.org 170 An Efficient Secure Biometric System with Non-Invertible Gabor Transform Radha Narayanan 1 and Kathikeyan Subramanian 2 1 Department of Computer Science, Karpagam University Coimbatore,641

More information

Fingerprint Image Enhancement Algorithm and Performance Evaluation

Fingerprint Image Enhancement Algorithm and Performance Evaluation Fingerprint Image Enhancement Algorithm and Performance Evaluation Naja M I, Rajesh R M Tech Student, College of Engineering, Perumon, Perinad, Kerala, India Project Manager, NEST GROUP, Techno Park, TVM,

More information

REINFORCED FINGERPRINT MATCHING METHOD FOR AUTOMATED FINGERPRINT IDENTIFICATION SYSTEM

REINFORCED FINGERPRINT MATCHING METHOD FOR AUTOMATED FINGERPRINT IDENTIFICATION SYSTEM REINFORCED FINGERPRINT MATCHING METHOD FOR AUTOMATED FINGERPRINT IDENTIFICATION SYSTEM 1 S.Asha, 2 T.Sabhanayagam 1 Lecturer, Department of Computer science and Engineering, Aarupadai veedu institute of

More information

Call for participation. FVC2004: Fingerprint Verification Competition 2004

Call for participation. FVC2004: Fingerprint Verification Competition 2004 Call for participation FVC2004: Fingerprint Verification Competition 2004 WEB SITE: http://bias.csr.unibo.it/fvc2004/ The Biometric System Lab (University of Bologna), the Pattern Recognition and Image

More information

Secure Fingerprint Matching with External Registration

Secure Fingerprint Matching with External Registration Secure Fingerprint Matching with External Registration James Reisman 1, Umut Uludag 2, and Arun Ross 3 1 Siemens Corporate Research, 755 College Road East, Princeton, NJ, 08540 james.reisman@siemens.com

More information

AN EFFICIENT METHOD FOR FINGERPRINT RECOGNITION FOR NOISY IMAGES

AN EFFICIENT METHOD FOR FINGERPRINT RECOGNITION FOR NOISY IMAGES International Journal of Computer Science and Communication Vol. 3, No. 1, January-June 2012, pp. 113-117 AN EFFICIENT METHOD FOR FINGERPRINT RECOGNITION FOR NOISY IMAGES Vijay V. Chaudhary 1 and S.R.

More information

One-Time Templates for Face Authentication

One-Time Templates for Face Authentication 27 International Conference on Convergence Information Technology One-Time Templates for Face Authentication Yongjin Lee, Yongki Lee, Yunsu Chung and Kiyoung Moon Biometrics Technology Research Team Electronics

More information

FILTERBANK-BASED FINGERPRINT MATCHING. Dinesh Kapoor(2005EET2920) Sachin Gajjar(2005EET3194) Himanshu Bhatnagar(2005EET3239)

FILTERBANK-BASED FINGERPRINT MATCHING. Dinesh Kapoor(2005EET2920) Sachin Gajjar(2005EET3194) Himanshu Bhatnagar(2005EET3239) FILTERBANK-BASED FINGERPRINT MATCHING Dinesh Kapoor(2005EET2920) Sachin Gajjar(2005EET3194) Himanshu Bhatnagar(2005EET3239) Papers Selected FINGERPRINT MATCHING USING MINUTIAE AND TEXTURE FEATURES By Anil

More information

Keywords Fingerprint enhancement, Gabor filter, Minutia extraction, Minutia matching, Fingerprint recognition. Bifurcation. Independent Ridge Lake

Keywords Fingerprint enhancement, Gabor filter, Minutia extraction, Minutia matching, Fingerprint recognition. Bifurcation. Independent Ridge Lake Volume 4, Issue 8, August 2014 ISSN: 2277 128X International Journal of Advanced Research in Computer Science and Software Engineering Research Paper Available online at: www.ijarcsse.com A novel approach

More information

A new approach to reference point location in fingerprint recognition

A new approach to reference point location in fingerprint recognition A new approach to reference point location in fingerprint recognition Piotr Porwik a) and Lukasz Wieclaw b) Institute of Informatics, Silesian University 41 200 Sosnowiec ul. Bedzinska 39, Poland a) porwik@us.edu.pl

More information

CHAPTER 6 EFFICIENT TECHNIQUE TOWARDS THE AVOIDANCE OF REPLAY ATTACK USING LOW DISTORTION TRANSFORM

CHAPTER 6 EFFICIENT TECHNIQUE TOWARDS THE AVOIDANCE OF REPLAY ATTACK USING LOW DISTORTION TRANSFORM 109 CHAPTER 6 EFFICIENT TECHNIQUE TOWARDS THE AVOIDANCE OF REPLAY ATTACK USING LOW DISTORTION TRANSFORM Security is considered to be the most critical factor in many applications. The main issues of such

More information

Verifying Fingerprint Match by Local Correlation Methods

Verifying Fingerprint Match by Local Correlation Methods Verifying Fingerprint Match by Local Correlation Methods Jiang Li, Sergey Tulyakov and Venu Govindaraju Abstract Most fingerprint matching algorithms are based on finding correspondences between minutiae

More information

Fingerprint matching using ridges

Fingerprint matching using ridges Fingerprint matching using ridges Jianjiang Feng a, *, Zhengyu Ouyang a, and Anni Cai a a Beijing University of Posts and Telecommunications, Box 113, Beijing, 100876, P. R. China *Corresponding author.

More information

A flexible biometrics remote user authentication scheme

A flexible biometrics remote user authentication scheme Computer Standards & Interfaces 27 (2004) 19 23 www.elsevier.com/locate/csi A flexible biometrics remote user authentication scheme Chu-Hsing Lin*, Yi-Yi Lai Department of Computer Science and Information

More information

Cancelable Biometrics

Cancelable Biometrics Cancelable Biometrics Chinthamreddy Premsai B.Tech(3rd Year) Computer Science & Engineering Indian Institute of Technology, Kharagpur email: premsaichinthamreddy@gmail.com Project Guide Dr. M.V.N.K Prasad

More information

Improving Latent Fingerprint Matching Performance by Orientation Field Estimation using Localized Dictionaries

Improving Latent Fingerprint Matching Performance by Orientation Field Estimation using Localized Dictionaries Available Online at www.ijcsmc.com International Journal of Computer Science and Mobile Computing A Monthly Journal of Computer Science and Information Technology IJCSMC, Vol. 3, Issue. 11, November 2014,

More information

Fingerprint Matching Using Minutiae Feature Hardikkumar V. Patel, Kalpesh Jadav

Fingerprint Matching Using Minutiae Feature Hardikkumar V. Patel, Kalpesh Jadav Fingerprint Matching Using Minutiae Feature Hardikkumar V. Patel, Kalpesh Jadav Abstract- Fingerprints have been used in identification of individuals for many years because of the famous fact that each

More information

Page Mapping Scheme to Support Secure File Deletion for NANDbased Block Devices

Page Mapping Scheme to Support Secure File Deletion for NANDbased Block Devices Page Mapping Scheme to Support Secure File Deletion for NANDbased Block Devices Ilhoon Shin Seoul National University of Science & Technology ilhoon.shin@snut.ac.kr Abstract As the amount of digitized

More information

Illumination-Robust Face Recognition based on Gabor Feature Face Intrinsic Identity PCA Model

Illumination-Robust Face Recognition based on Gabor Feature Face Intrinsic Identity PCA Model Illumination-Robust Face Recognition based on Gabor Feature Face Intrinsic Identity PCA Model TAE IN SEOL*, SUN-TAE CHUNG*, SUNHO KI**, SEONGWON CHO**, YUN-KWANG HONG*** *School of Electronic Engineering

More information

Fast and Robust Projective Matching for Fingerprints using Geometric Hashing

Fast and Robust Projective Matching for Fingerprints using Geometric Hashing Fast and Robust Projective Matching for Fingerprints using Geometric Hashing Rintu Boro Sumantra Dutta Roy Department of Electrical Engineering, IIT Bombay, Powai, Mumbai - 400 076, INDIA {rintu, sumantra}@ee.iitb.ac.in

More information

Comparison of ROC-based and likelihood methods for fingerprint verification

Comparison of ROC-based and likelihood methods for fingerprint verification Comparison of ROC-based and likelihood methods for fingerprint verification Sargur Srihari, Harish Srinivasan, Matthew Beal, Prasad Phatak and Gang Fang Department of Computer Science and Engineering University

More information

Fingerprint Deformation Models Using Minutiae Locations and Orientations

Fingerprint Deformation Models Using Minutiae Locations and Orientations Fingerprint Deformation Models Using Minutiae Locations and Orientations Yi Chen, Sarat Dass, Arun Ross, and Anil Jain Department of Computer Science and Engineering Michigan State University East Lansing,

More information

A Fast Personal Palm print Authentication based on 3D-Multi Wavelet Transformation

A Fast Personal Palm print Authentication based on 3D-Multi Wavelet Transformation A Fast Personal Palm print Authentication based on 3D-Multi Wavelet Transformation * A. H. M. Al-Helali, * W. A. Mahmmoud, and * H. A. Ali * Al- Isra Private University Email: adnan_hadi@yahoo.com Abstract:

More information

Rotation Invariant Finger Vein Recognition *

Rotation Invariant Finger Vein Recognition * Rotation Invariant Finger Vein Recognition * Shaohua Pang, Yilong Yin **, Gongping Yang, and Yanan Li School of Computer Science and Technology, Shandong University, Jinan, China pangshaohua11271987@126.com,

More information

A Hybrid Approach for Generating Secure and Discriminating Face Template Yi C. Feng, Pong C. Yuen, Member, IEEE, and Anil K. Jain, Fellow, IEEE

A Hybrid Approach for Generating Secure and Discriminating Face Template Yi C. Feng, Pong C. Yuen, Member, IEEE, and Anil K. Jain, Fellow, IEEE IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, VOL. 5, NO. 1, MARCH 2010 103 A Hybrid Approach for Generating Secure and Discriminating Face Template Yi C. Feng, Pong C. Yuen, Member, IEEE, and

More information

A NOVEL APPROACH FOR GENERATING FACE TEMPLATE USING BDA

A NOVEL APPROACH FOR GENERATING FACE TEMPLATE USING BDA A NOVEL APPROACH FOR GENERATING FACE TEMPLATE USING BDA Shraddha S. Shinde 1 and Prof. Anagha P. Khedkar 2 1 P.G. Student, Department of Computer Engineering, MCERC, Nashik (M.S.), India. shraddhashinde@gmail.com

More information

Reconstructing Ridge Frequency Map from Minutiae Template of Fingerprints

Reconstructing Ridge Frequency Map from Minutiae Template of Fingerprints Reconstructing Ridge Frequency Map from Minutiae Template of Fingerprints Wei Tang, Yukun Liu College of Measurement & Control Technology and Communication Engineering Harbin University of Science and

More information

DEFORMABLE MATCHING OF HAND SHAPES FOR USER VERIFICATION. Ani1 K. Jain and Nicolae Duta

DEFORMABLE MATCHING OF HAND SHAPES FOR USER VERIFICATION. Ani1 K. Jain and Nicolae Duta DEFORMABLE MATCHING OF HAND SHAPES FOR USER VERIFICATION Ani1 K. Jain and Nicolae Duta Department of Computer Science and Engineering Michigan State University, East Lansing, MI 48824-1026, USA E-mail:

More information

Template Protection and its Implementation in 3D Face Recognition Systems

Template Protection and its Implementation in 3D Face Recognition Systems Template Protection and its Implementation in 3D Face Recognition Systems Xuebing Zhou Fraunhofer IGD, Fraunhoferstr. 5, 64283 Darmstadt, Germany E-mail: xuebing.zhou@igd.fhg.de ABSTRACT As biometric recognition

More information

Ground truth and evaluation for latent fingerprint matching

Ground truth and evaluation for latent fingerprint matching Ground truth and evaluation for latent fingerprint matching Anna Mikaelyan and Josef Bigun Halmstad University SE-0118 Halmstad {anna.mikaelyan,josef.bigun}@hh.se Abstract In forensic fingerprint studies

More information

International Journal of Advanced Research in Computer Science and Software Engineering

International Journal of Advanced Research in Computer Science and Software Engineering ISSN: 2277 128X International Journal of Advanced Research in Computer Science and Software Engineering Research Paper Available online at: Fingerprint Recognition using Robust Local Features Madhuri and

More information

Reliability Measure of 2D-PAGE Spot Matching using Multiple Graphs

Reliability Measure of 2D-PAGE Spot Matching using Multiple Graphs Reliability Measure of 2D-PAGE Spot Matching using Multiple Graphs Dae-Seong Jeoune 1, Chan-Myeong Han 2, Yun-Kyoo Ryoo 3, Sung-Woo Han 4, Hwi-Won Kim 5, Wookhyun Kim 6, and Young-Woo Yoon 6 1 Department

More information

EE368 Project: Visual Code Marker Detection

EE368 Project: Visual Code Marker Detection EE368 Project: Visual Code Marker Detection Kahye Song Group Number: 42 Email: kahye@stanford.edu Abstract A visual marker detection algorithm has been implemented and tested with twelve training images.

More information

Filterbank-Based Fingerprint Matching. Multimedia Systems Project. Niveditha Amarnath Samir Shah

Filterbank-Based Fingerprint Matching. Multimedia Systems Project. Niveditha Amarnath Samir Shah Filterbank-Based Fingerprint Matching Multimedia Systems Project Niveditha Amarnath Samir Shah Presentation overview Introduction Background Algorithm Limitations and Improvements Conclusions and future

More information

Incorporating Image Quality in Multi-Algorithm Fingerprint Verification

Incorporating Image Quality in Multi-Algorithm Fingerprint Verification Incorporating Image Quality in Multi-Algorithm Fingerprint Verification Julian Fierrez-Aguilar 1, Yi Chen 2, Javier Ortega-Garcia 1, and Anil K. Jain 2 1 ATVS, Escuela Politecnica Superior, Universidad

More information

OPTIMIZED DUAL FINGERPRINT MECHANISM FOR PRIVACY PROTECTION

OPTIMIZED DUAL FINGERPRINT MECHANISM FOR PRIVACY PROTECTION OPTIMIZED DUAL FINGERPRINT MECHANISM FOR PRIVACY PROTECTION 1 Sanjyoti Lakhe, Student of ME (CSE), Government College of Engineering,Aurangabad, Dr.Babasaheb Ambedkar Marathwada University, Aurangabad.

More information

The Design of Fingerprint Biometric Authentication on Smart Card for

The Design of Fingerprint Biometric Authentication on Smart Card for The Design of Fingerprint Biometric Authentication on Smart Card for PULAPOT Main Entrance System Computer Science Department, Faculty of Technology Science and Defence Universiti Pertahanan Nasional Malaysia

More information

AN EFFICIENT BINARIZATION TECHNIQUE FOR FINGERPRINT IMAGES S. B. SRIDEVI M.Tech., Department of ECE

AN EFFICIENT BINARIZATION TECHNIQUE FOR FINGERPRINT IMAGES S. B. SRIDEVI M.Tech., Department of ECE AN EFFICIENT BINARIZATION TECHNIQUE FOR FINGERPRINT IMAGES S. B. SRIDEVI M.Tech., Department of ECE sbsridevi89@gmail.com 287 ABSTRACT Fingerprint identification is the most prominent method of biometric

More information

FVC2004: Third Fingerprint Verification Competition

FVC2004: Third Fingerprint Verification Competition FVC2004: Third Fingerprint Verification Competition D. Maio 1, D. Maltoni 1, R. Cappelli 1, J.L. Wayman 2, A.K. Jain 3 1 Biometric System Lab - DEIS, University of Bologna, via Sacchi 3, 47023 Cesena -

More information

CSE / 60537: Biometrics

CSE / 60537: Biometrics CSE 40537 / 60537: Biometrics * ** * * ** * * Template Protection 3 62 The Course Instructor Feedback (CIF) survey window for biometrics opens tomorrow December 1st - 13th https://cif.nd.edu/ How'm I doin'?

More information

Preprocessing of a Fingerprint Image Captured with a Mobile Camera

Preprocessing of a Fingerprint Image Captured with a Mobile Camera Preprocessing of a Fingerprint Image Captured with a Mobile Camera Chulhan Lee 1, Sanghoon Lee 1,JaihieKim 1, and Sung-Jae Kim 2 1 Biometrics Engineering Research Center, Department of Electrical and Electronic

More information

DIGITAL IMAGE PROCESSING APPROACH TO FINGERPRINT AUTHENTICATION

DIGITAL IMAGE PROCESSING APPROACH TO FINGERPRINT AUTHENTICATION DAAAM INTERNATIONAL SCIENTIFIC BOOK 2012 pp. 517-526 CHAPTER 43 DIGITAL IMAGE PROCESSING APPROACH TO FINGERPRINT AUTHENTICATION RAKUN, J.; BERK, P.; STAJNKO, D.; OCEPEK, M. & LAKOTA, M. Abstract: In this

More information

Efficient Rectification of Malformation Fingerprints

Efficient Rectification of Malformation Fingerprints Efficient Rectification of Malformation Fingerprints Ms.Sarita Singh MCA 3 rd Year, II Sem, CMR College of Engineering & Technology, Hyderabad. ABSTRACT: Elastic distortion of fingerprints is one of the

More information

Hiding Fingerprint Minutiae in Images

Hiding Fingerprint Minutiae in Images Hiding Fingerprint Minutiae in Images Anil K. Jain and Umut Uludag Computer Science and Engineering Department, Michigan State University 3115 Engineering Building, East Lansing, MI, 48824, USA {jain,

More information

A Study on the Consistency of Features for On-line Signature Verification

A Study on the Consistency of Features for On-line Signature Verification A Study on the Consistency of Features for On-line Signature Verification Center for Unified Biometrics and Sensors State University of New York at Buffalo Amherst, NY 14260 {hlei,govind}@cse.buffalo.edu

More information

Comparison of fingerprint enhancement techniques through Mean Square Error and Peak-Signal to Noise Ratio

Comparison of fingerprint enhancement techniques through Mean Square Error and Peak-Signal to Noise Ratio Comparison of fingerprint enhancement techniques through Mean Square Error and Peak-Signal to Noise Ratio M. M. Kazi A. V. Mane R. R. Manza, K. V. Kale, Professor and Head, Abstract In the fingerprint

More information

Fingerprint Recognition System for Low Quality Images

Fingerprint Recognition System for Low Quality Images Fingerprint Recognition System for Low Quality Images Zin Mar Win and Myint Myint Sein University of Computer Studies, Yangon, Myanmar zmwucsy@gmail.com Department of Research and Development University

More information

Published by: PIONEER RESEARCH & DEVELOPMENT GROUP (www.prdg.org) 1

Published by: PIONEER RESEARCH & DEVELOPMENT GROUP (www.prdg.org) 1 Minutiae Points Extraction using Biometric Fingerprint- Enhancement Vishal Wagh 1, Shefali Sonavane 2 1 Computer Science and Engineering Department, Walchand College of Engineering, Sangli, Maharashtra-416415,

More information

Fingerprint verification by decision-level fusion of optical and capacitive sensors

Fingerprint verification by decision-level fusion of optical and capacitive sensors Fingerprint verification by decision-level fusion of optical and capacitive sensors Gian Luca Marcialis and Fabio Roli Department of Electrical and Electronic Engineering University of Cagliari Piazza

More information

Minutiae Based Fingerprint Authentication System

Minutiae Based Fingerprint Authentication System Minutiae Based Fingerprint Authentication System Laya K Roy Student, Department of Computer Science and Engineering Jyothi Engineering College, Thrissur, India Abstract: Fingerprint is the most promising

More information

Fingerprint matching using ridges

Fingerprint matching using ridges Pattern Recognition 39 (2006) 2131 2140 www.elsevier.com/locate/patcog Fingerprint matching using ridges Jianjiang Feng, Zhengyu Ouyang, Anni Cai School of Telecommunication Engineering, Beijing University

More information

FINGERPRINT RECOGNITION SYSTEM USING SUPPORT VECTOR MACHINE AND NEURAL NETWORK

FINGERPRINT RECOGNITION SYSTEM USING SUPPORT VECTOR MACHINE AND NEURAL NETWORK International Journal of Computer Science Engineering and Information Technology Research (IJCSEITR) ISSN(P): 2249-6831; ISSN(E): 2249-7943 Vol. 4, Issue 1, Feb 2014, 103-110 TJPRC Pvt. Ltd. FINGERPRINT

More information

Fingerprint Recognition Using Gabor Filter And Frequency Domain Filtering

Fingerprint Recognition Using Gabor Filter And Frequency Domain Filtering IOSR Journal of Electronics and Communication Engineering (IOSRJECE) ISSN : 2278-2834 Volume 2, Issue 6 (Sep-Oct 2012), PP 17-21 Fingerprint Recognition Using Gabor Filter And Frequency Domain Filtering

More information

On-Line Fingerprint Verification

On-Line Fingerprint Verification On-Line Fingerprint Verification Ani1 Jain and Lin Hong Pattern Recognition and Image Processing Laboratory Department of Computer Science Michigan State University East Lansing, MI 48824, USA {j ain,honglin}

More information

A FINGER PRINT RECOGNISER USING FUZZY EVOLUTIONARY PROGRAMMING

A FINGER PRINT RECOGNISER USING FUZZY EVOLUTIONARY PROGRAMMING A FINGER PRINT RECOGNISER USING FUZZY EVOLUTIONARY PROGRAMMING Author1: Author2: K.Raghu Ram K.Krishna Chaitanya 4 th E.C.E 4 th E.C.E raghuram.kolipaka@gmail.com chaitu_kolluri@yahoo.com Newton s Institute

More information

Fingerprint Identification Using SIFT-Based Minutia Descriptors and Improved All Descriptor-Pair Matching

Fingerprint Identification Using SIFT-Based Minutia Descriptors and Improved All Descriptor-Pair Matching Sensors 2013, 13, 3142-3156; doi:10.3390/s130303142 Article OPEN ACCESS sensors ISSN 1424-8220 www.mdpi.com/journal/sensors Fingerprint Identification Using SIFT-Based Minutia Descriptors and Improved

More information

Abstract -Fingerprints are the most widely. Keywords:fingerprint; ridge pattern; biometric;

Abstract -Fingerprints are the most widely. Keywords:fingerprint; ridge pattern; biometric; Analysis Of Finger Print Detection Techniques Prof. Trupti K. Wable *1(Assistant professor of Department of Electronics & Telecommunication, SVIT Nasik, India) trupti.wable@pravara.in*1 Abstract -Fingerprints

More information

Genetic Model Optimization for Hausdorff Distance-Based Face Localization

Genetic Model Optimization for Hausdorff Distance-Based Face Localization c In Proc. International ECCV 2002 Workshop on Biometric Authentication, Springer, Lecture Notes in Computer Science, LNCS-2359, pp. 103 111, Copenhagen, Denmark, June 2002. Genetic Model Optimization

More information

Lane Detection using Fuzzy C-Means Clustering

Lane Detection using Fuzzy C-Means Clustering Lane Detection using Fuzzy C-Means Clustering Kwang-Baek Kim, Doo Heon Song 2, Jae-Hyun Cho 3 Dept. of Computer Engineering, Silla University, Busan, Korea 2 Dept. of Computer Games, Yong-in SongDam University,

More information

A Fast Minutiae-Based Fingerprint Recognition System Haiyun Xu, Raymond N. J. Veldhuis, Tom A. M. Kevenaar, and Ton A. H. M.

A Fast Minutiae-Based Fingerprint Recognition System Haiyun Xu, Raymond N. J. Veldhuis, Tom A. M. Kevenaar, and Ton A. H. M. 418 IEEE SYSTEMS JOURNAL, VOL. 3, NO. 4, DECEMBER 2009 A Fast Minutiae-Based Fingerprint Recognition System Haiyun Xu, Raymond N. J. Veldhuis, Tom A. M. Kevenaar, and Ton A. H. M. Akkermans Abstract The

More information

Fingerprint Ridge Orientation Estimation Using A Modified Canny Edge Detection Mask

Fingerprint Ridge Orientation Estimation Using A Modified Canny Edge Detection Mask Fingerprint Ridge Orientation Estimation Using A Modified Canny Edge Detection Mask Laurice Phillips PhD student laurice.phillips@utt.edu.tt Margaret Bernard Senior Lecturer and Head of Department Margaret.Bernard@sta.uwi.edu

More information

Robust biometric image watermarking for fingerprint and face template protection

Robust biometric image watermarking for fingerprint and face template protection Robust biometric image watermarking for fingerprint and face template protection Mayank Vatsa 1, Richa Singh 1, Afzel Noore 1a),MaxM.Houck 2, and Keith Morris 2 1 West Virginia University, Morgantown,

More information

Feature-level Fusion for Effective Palmprint Authentication

Feature-level Fusion for Effective Palmprint Authentication Feature-level Fusion for Effective Palmprint Authentication Adams Wai-Kin Kong 1, 2 and David Zhang 1 1 Biometric Research Center, Department of Computing The Hong Kong Polytechnic University, Kowloon,

More information

Focal Point Detection Based on Half Concentric Lens Model for Singular Point Extraction in Fingerprint

Focal Point Detection Based on Half Concentric Lens Model for Singular Point Extraction in Fingerprint Focal Point Detection Based on Half Concentric Lens Model for Singular Point Extraction in Fingerprint Natthawat Boonchaiseree and Vutipong Areekul Kasetsart Signal & Image Processing Laboratory (KSIP

More information

Providing Authentication by Merging Minutiae Template.

Providing Authentication by Merging Minutiae Template. Providing Authentication by Merging Minutiae Template. Priya Raul *1, Sayali Surve* 2, Sushma Gilbile *3, Jasmine Hebbalkar *4, Prof J.L. Chaudhari #5, *B.E Students, #Assistant Professor, Department of

More information

Fingerprint Ridge Distance Estimation: Algorithms and the Performance*

Fingerprint Ridge Distance Estimation: Algorithms and the Performance* Fingerprint Ridge Distance Estimation: Algorithms and the Performance* Xiaosi Zhan, Zhaocai Sun, Yilong Yin, and Yayun Chu Computer Department, Fuyan Normal College, 3603, Fuyang, China xiaoszhan@63.net,

More information

Secure and Private Identification through Biometric Systems

Secure and Private Identification through Biometric Systems Secure and Private Identification through Biometric Systems 1 Keshav Rawat, 2 Dr. Chandra Kant 1 Assistant Professor, Deptt. of Computer Science & Informatics, C.U. Himachal Pradesh Dharamshala 2 Assistant

More information

Biometrics Technology: Hand Geometry

Biometrics Technology: Hand Geometry Biometrics Technology: Hand Geometry References: [H1] Gonzeilez, S., Travieso, C.M., Alonso, J.B., and M.A. Ferrer, Automatic biometric identification system by hand geometry, Proceedings of IEEE the 37th

More information

Computer Vision and Image Understanding

Computer Vision and Image Understanding Computer Vision and Image Understanding 113 (2009) 979 992 Contents lists available at ScienceDirect Computer Vision and Image Understanding journal homepage: www.elsevier.com/locate/cviu Minutiae-based

More information

SIFT: SCALE INVARIANT FEATURE TRANSFORM SURF: SPEEDED UP ROBUST FEATURES BASHAR ALSADIK EOS DEPT. TOPMAP M13 3D GEOINFORMATION FROM IMAGES 2014

SIFT: SCALE INVARIANT FEATURE TRANSFORM SURF: SPEEDED UP ROBUST FEATURES BASHAR ALSADIK EOS DEPT. TOPMAP M13 3D GEOINFORMATION FROM IMAGES 2014 SIFT: SCALE INVARIANT FEATURE TRANSFORM SURF: SPEEDED UP ROBUST FEATURES BASHAR ALSADIK EOS DEPT. TOPMAP M13 3D GEOINFORMATION FROM IMAGES 2014 SIFT SIFT: Scale Invariant Feature Transform; transform image

More information

Image Enhancement Techniques for Fingerprint Identification

Image Enhancement Techniques for Fingerprint Identification March 2013 1 Image Enhancement Techniques for Fingerprint Identification Pankaj Deshmukh, Siraj Pathan, Riyaz Pathan Abstract The aim of this paper is to propose a new method in fingerprint enhancement

More information

Fingerprint Indexing using Minutiae and Pore Features

Fingerprint Indexing using Minutiae and Pore Features Fingerprint Indexing using Minutiae and Pore Features R. Singh 1, M. Vatsa 1, and A. Noore 2 1 IIIT Delhi, India, {rsingh, mayank}iiitd.ac.in 2 West Virginia University, Morgantown, USA, afzel.noore@mail.wvu.edu

More information

Published by: PIONEER RESEARCH & DEVELOPMENT GROUP (www.prdg.org) 1

Published by: PIONEER RESEARCH & DEVELOPMENT GROUP (www.prdg.org) 1 Enhancing Security in Identity Documents Using QR Code RevathiM K 1, Annapandi P 2 and Ramya K P 3 1 Information Technology, Dr.Sivanthi Aditanar College of Engineering, Tiruchendur, Tamilnadu628215, India

More information

FINGERPRINT RECOGNITION BASED ON SPECTRAL FEATURE EXTRACTION

FINGERPRINT RECOGNITION BASED ON SPECTRAL FEATURE EXTRACTION FINGERPRINT RECOGNITION BASED ON SPECTRAL FEATURE EXTRACTION Nadder Hamdy, Magdy Saeb 2, Ramy Zewail, and Ahmed Seif Arab Academy for Science, Technology & Maritime Transport School of Engineering,. Electronics

More information