Fingerprint Matching Approach Based on Bifurcation Minutiae

JOURNAL OF INFORMATION AND COMMUNICATION TECHNOLOGIES, VOLUME 2, ISSUE 5, MAY 2012 33

Fingerprint Matching Approach Based on Bifurcation Minutiae Dr. Manal Abdullah, Mona Alkhozae and Mashaiel Alsowaiel Faculty of Computing and Information Technology (FCIT), Department of Computer Sciences King Abdulaziz University, Jeddah, KSA

Abstract— Abstract — Fingerprints are the most widely used biometric feature for user identification and verification in the field of biometric identification and security. Minutiae based approach involves extraction of minutiae points from the sample fingerprint images and performing fingerprint matching based on the number of minutiae pairings among two fingerprints in question and generates a matching percentage. Matching techniques still suffer from distortion problems due to poor quality impressions which lead to difficulties in forming a match among multiple impressions acquired from same fingertip. The minutiae from the ridge ending are mostly unreliable due to the poor quality impressions. In this paper we use only the bifurcation minutiae points in the alignment matching stage instead of using it with ending minutiae points. Also we use the bifurcation minutiae points with the ending minutiae points in the match stage. We test the effect of increasing the number of training sets on the performance of our matching technique where it highly reduces FRR when using two templates. The experiment results ensure performance improvements of the enhanced minutiae based approach by taking only the bifurcation minutiae points in the alignment stage sets. The research is coded using MATLAB. Index Terms—Fingerprints, Terms —Fingerprints, biometric, Minutiae based approach, image enhancement, Image binarization, Image segmentation, Image thinning, Minutiae marking, False Minutiae Removal.

——————————

1. INTRODUCTION

M

u

—————————— fingerprint matching based on the number of minutiae pairings among two fingerprints in question and generates a matching percentage for the two fingerprints similarity. An alignment-based match algorithm partially includes two successive stages, one is for alignment stage and the second is the match stage. Because the minutiae from the ridge ending are mostly unreliable due the poor quality impressions so we discard them from the sets in the alignment stage and calculate the similarity of the two ridges associated with two referenced bifurcation minutiae points and if the similarity is larger than certain predefined threshold value, the algorithm transforms each set of minutiae to a new coordination system whose origin is at the referenced point and whose x-axis is coincident with the direction of the referenced point. Match stage use the elastic match algorithm to count the matched minutiae pairs by assuming two minutiae having almost the same position and direction are identical.

ore than a century has passed since Alphonse Bertillon first conceived and then industriously practiced the idea of using body measurements for solving crimes. Just as his idea was gaining popularity, it faded into relative obscurity by a far more significant and practical discovery of the distinctiveness of the human fingerprints. In 1893, the Home Ministry Office, UK, accepted that no two individuals have the same fingerprints. Soon after that, many major law enforcement departments saw potential of fingerprints in identifying repeat offenders who used an alias. They successfully applied the art of fingerprint recognition. The manual method of fingerprint indexing (based on the Henry system of classification) resulted in a highly skewed distribution of fingerprints into bins (types): most fingerprints fell into a few bins and this did not improve 2. FINGERPRINT RECOGNITION the search efficiency. Fingerprint training procedures were An important issue in designing a practical biometric time-intensive and slow. Furthermore, demands imposed system is to determine how an individual is going to be by the painstaking attention needed to visually compare recognized. Depending on the application context, a two fingerprints of varied qualities, tedium of the biometric system may be called either a verification system monotonous nature of the work, and increasing workloads or an identification system [1]: due to a higher demand on fingerprint recognition services. A verification system authenticates a person’s The efforts led to the development of Automatic identity by comparing the captured biometric Fingerprint Identification Systems (AFIS). More recently, characteristic with his/her previously captured however, increasing concerns about security and identity (enrolled) biometric reference template pre-stored in fraud have created a growing need for fingerprint and the system. It conducts one-to-one comparison. other biometric technologies for person recognition in a An identification system recognizes an individual by searching the entire enrolment template database for large number of non-forensic applications [1]. a match. It conducts one-to-many comparisons. In this paper we use minutiae based fingerprint recognition. It involves extraction of minutiae points from the sample fingerprint images and then performing •

•

© 2012 JICT www.jict.co.uk


The block diagrams of fingerprint verification and identification systems are depicted in Figure 1 [1].

invariant and discriminating features are extracted from fingerprint image. Fingerprint classification can be viewed as coarse level matching. Since each match takes some amount of time, the maximum size of database is limited. A solution to this problem is to classify fingerprint, and first matching the fingerprints that use small features. Based on the observations of how human fingerprint experts perform fingerprint recognition, three major problems in designing AFISs were identified and investigated: digital fingerprint acquisition, local ridge characteristic extraction, and ridge characteristic pattern matching [1]. There are three specific classes for all fingerprints based upon their general visual pattern. These are: loops, whorls, and arches. Approximately 60% of the total population have loops, 35% have whorls, and 5% have arches. The three major groups are also subcategorized based upon smaller differences existing between the patterns within the specific group. These subcategories are summarized in table 1 [2]: TABLE 1: FINGERPRINT PATTERN TYPES I. ARCH

Figure 1.Fingerprint 1.Fingerprint Enrolment, verification, and identification processes.

The enrolment, verification, and identification processes involved the following system modules [1]: Capture: a digital representation of biometric characteristic needs to be sensed and captured by a biometric sensor, such as a fingerprint scanner. Feature extraction: the digital representation (sample) is usually further processed by a feature extractor to generate a feature set to facilitate matching or comparison. Template creation: the template creation module organizes one or more feature sets into an enrolment template that will be saved in some persistent storage. Pre-selection and matching: the pre-selection (or filtering) stage is used in an identification system when the number of enrolled templates is large. Its role is to reduce the effective size of the template database. The matching (or comparison) stage takes a feature set and an enrolment template as inputs and computes the similarity between them in terms of a matching score. The matching score is compared to a system threshold to make the final decision. Data storage: is devoted to storing templates and other demographic information about the user. •

•

•

•

•

2.1 Fingerprint Identification:

Fingerprint identification is one of the most important biometric technologies since the unchangeable fingerprints during human life span, and the uniqueness of each individual’s fingerprints. A fingerprint is the pattern of ridges and valleys on the surface of a fingertip. The ridges are the dark areas of the fingerprint and the valleys are the white areas that exist between the ridges. The uniqueness of the fingerprint can be identified by the characteristics and relationships of bifurcations and endings in ridge or valley. In order to compare two fingerprints, a set of

a) Plain arch b) Tented arch

II. LOOP a) Radial loop b) Ulnar loop

III. WHORL a) Plain whorl b) Central pocket whorl c) Double loop d) Accidental whorl

Based on the mode of acquisition, a fingerprint image may be classified as off-line or live scan. An off-line image is typically obtained by smearing ink on the fingertip and creating an inked impression of the fingertip on paper. The inked impression is then digitized by scanning the paper using an optical scanner or a high-quality video camera. A live-scan image, on the other hand, is acquired by sensing the tip of the finger directly, using a sensor that is capable of digitizing the fingerprint on contact. Particular kind of off-line images, extremely important in forensic applications, are the so-called latent fingerprints found at crime scenes. The oily nature of the skin results in the impression of a fingerprint being deposited on a surface that is touched by a finger. [1]. 2.2 Fingerprint Classification

Fingerprint classification involving 6 classes with critical points in a fingerprint called core and delta marked as circles and triangles given in a figure 2. Many classifications are given to patterns that can arise in the ridges of a fingerprint (shown in figure3). These points are called the minutiae of the fingerprint. The most commonly used minutiae in current fingerprint recognition technologies are ridge endings and bifurcations, because they can be easily detected by only looking at points that surround them (Bifurcation is the location where a ridge divides into two separate ridges). A good quality fingerprint contains 30 – 80 minutiae points [3].



collecting fingerprint samples, there are only few benchmark databases database s available. Moreover, various types of fingerprints are required to measure how robust the system is in various environments. Some of fingerprint databases are listed in table2 [5]. We use the database provided by FVC2002 DB2 (Fingerprint Verification Competition 2002) the fingerprint were captured using optical sensor and the image size is296*560pixel and its resolution is 569 dpi, we choose seven persons and 3 images for each one. TABLE 2: FINGERPRINT DATABASES Figure 2 Fingerprint classifications.

A fingerprint is an impression of the epidermal ridges of a human fingertip. A hierarchy of three levels of features, namely, Level 1 (pattern), Level 2 (minutiae points) and Level 3 (pores and ridge shape) are used for recognition purposes.

Level 1 feature refer to the overall pattern shape of the unknown fingerprint. Level 2 features refers to specific friction ridge paths like ridge endings, lakes, islands, bifurcations, scars, incipient ridges, and flexion creases. Level 3 details refer to the intrinsic detail present in a developed fingerprint pores, ridge units, edge detail, scars, etc. The three levels are depicted in figure 3 [3].

Figure 3 Fingerprint Features.

Minutiae features refer to micro-node characters which are also categorized five kinds: ending, bifurcation, liprounding, double bifurcations and bridge, shown in Figure 4[4].

Figure 4.Category of minutiae fingerprints feature.

2.3 Fingerprint Databases:

Fingerprint identification requires that the fingerprint match a pre-recorded fingerprint template stored in the fingerprint database. Constructing a fingerprint database is important to evaluate the performance of automatic fingerprint recognition systems. Because of the difficulty in

Fingerprint Database CASIA Fingerprint Image Database [6]:

Features It contains 20,000 fingerprint images of 4,000 fingers from 500 subjects. The volunteers include graduate students, workers, waiters, etc. Each volunteer contributed 40 fingerprint images of his eight fingers (left and right thumb/second/third/fourth finger), i.e. 5 images per finger. All fingerprint images are 8 bit gray-level BMP files and the image resolution is 328*356 The NIST It is publicly available and fingerprint database contains 8 bit grayscale [7]: fingerprint image pairs each print is 512 X 512 pixels with 32 rows of white space at the bottom of the print. The database is being distributed for use in the development and testing of automated fingerprint classification systems on a common set of images. Fingerprint FVC2006 [8]: There are four Verification distinct databases, DB1, DB2, Competition DB3 and DB4. Each database is databases (FVC) : 150 fingers wide and 12 samples per finger in depth (i.e., it consists of 1800 fingerprint images). The image format is BMP, 256 gray-levels, uncompressed. FVC2004 [9]: There are Four different databases (DB1, DB2, DB3 and DB4) and for each database a total of 120 fingers and 12 impressions per finger (1440 impressions) were gathered. The size of each database to be used in the test was established as 110 fingers wide (w) and 8 impressions per finger deep (d) (880 fingerprints in all). FVC2002 [10]: There are Four different databases (DB1, DB2, DB3 and DB4) and each database is 110 fingers wide (w) and 8 impressions per finger deep (d) (880 fingerprints in all).



FVC2000 [11]: There are Four different databases (DB1, DB2, DB3 and DB4) and each database is 110 fingers wide (w) and 8 impressions per finger deep (d) (880 fingerprints in all).

interest are obtained, a new modified thinning method using slide neighbourhood processing is applied to clarify the endpoints and the bifurcations in each specific pixel. The proposed methods here going through these stages are Normalization, Binarization, Quality markup and Thinning.

4. RESEARCH METHODOLOGY: 3. RELATED WORK: A major approach for fingerprint recognition today is to extract minutiae from fingerprint images and to perform fingerprint matching based on the number of corresponding minutiae pairings. One of the most difficult problems in fingerprint recognition has been that the recognition performance is significantly influenced by fingertip surface condition, which may vary depending on environmental or personal causes. Addressing this problem, Koichi et al [13] proposed a fingerprint recognition algorithm using phase-based image matching. The use of phase components in 2D discrete Fourier transforms of fingerprint images makes possible to achieve highly robust fingerprint recognition for low-quality fingerprints. Experimental evaluation using a set of fingerprint images captured from fingertips with difficult conditions (e.g., dry fingertips, rough fingertips, allergicskin fingertips) demonstrates an efficient recognition performance of the proposed algorithm compared with a typical minutiae-based algorithm. Afsar et al [14] presents the implementation of a minutiae based approach to fingerprint identification and verification and serves as a review of the different techniques used in various steps in the development of minutiae based AFIS. The technique conferred in this research is based on the extraction of minutiae from the thinned, binarized and segmented version of a fingerprint image. The system uses fingerprint classification for indexing during fingerprint matching which greatly enhances the performance of the matching algorithm. Good results (~92% accuracy) were obtained using the FVC2000 fingerprint databases. Khalid et al [15] provide an approach for matching fingerprints that combines local structure and moment’s features for matching regions between the input and template images. Their matching algorithm includes two main steps. At first, searching the matched regions based on local structures and invariant moment features, and then the algorithm is repeated until finding all the matched regions. The proposed method performs better accuracy than the traditional matching based only on local structure features. The experimental test is conducted with database FVC2002 DB. Sepasian et al [16] proposed a novel method for enhancement of fingerprint image by means of eliminating the artificial induced boundaries, precise background omission; avoid over saturation of the image in homogeneous areas, with no intermediate filtering and computational complexity. In order to eliminate the induced boundaries, CLAHE (contrast limited adaptive histogram equalization) technique is employed. ‘Clip Limit’ is then applied in order to avoid over saturation of the image in homogeneous areas. Subsequently, the image is disintegrated into an array of distinct blocks and the discrimination of the blocks is obtained by computing the standard deviation of the matrix elements to remove the image background. Once the boundaries for the region of

In our research we use minutiae based fingerprint recognition where we use only the bifurcation minutiae points in the alignment matching stage instead of using it with ending minutiae points. Also we use the bifurcation minutiae points with the ending minutiae points in the match stage. We conduct our experiment on the database provided by FVC2002 DB2 (Fingerprint Verification Competition 2002) and chooses even persons with 3 images for each one. One important factor that affects recognition performance is the number of images in the training set. So, before applying our algorithm, we test the effect of increasing the number of training sets on the performance of our matching technique. 4.1 Fingerprint extraction algorithm:

We use minutiae based on fingerprint recognition to extract the fingerprint minutiae. It consists of enhancement, binarization, segmentation, thinning, minutiae marking, false minutiae removal (Remarks) [12]. A. Fingerprint Image Enhancement :

Fingerprint Image enhancement is to make the image clearer for easy further operations. Since the fingerprint images acquired from sensors or other media are not assured with perfect quality, those enhancement methods, for increasing the contrast between ridges and valleys and for connecting the false broken points of ridges due to insufficient amount of ink, are very useful for keep a higher accuracy to fingerprint recognition. We use Histogram Equalization (HE) method to enhance the contrast of images by transforming its intensity values for reducing the influence of the noise in fingerprint images [12]. Fingerprint Image Binarization : Fingerprint Image Binarization is to transform the 8-bit gray fingerprint image to a 1-bit image with 0-value for ridges and 1-value for valleys. A locally adaptive binarization method is performed to binarize the fingerprint image [12]. Such a named method comes from the mechanism of transforming a pixel value to 1 if the value is larger than the mean intensity value of the current block (16x16) to which the pixel belongs. B.

Fingerprint Image Segmentation: Region of Interest (ROI) is useful to be recognized for each fingerprint image. The image area without effective ridges and valleys is first discarded since it only holds background information. Then the bound of o f the remaining effective area is sketched out since the minutiae in the bound region are confusing with those spurious minutiae that are generated when the ridges are out of the sensor. Figure 5 shows the interest fingerprint image area and it's bound. C.



Figure 8: Bifurcation and ending ending minutiae marking

Figure 5: After OPEN operation (left), ROI + Bound (right).

F. False Minutiae Removal (Remarks) :

Three Processes for marking are involved: Process 1 : if distance between centre of bifurcation and

D. Fingerprint Ridge Thinning:

Ridge thinning is to eliminate the redundant pixels of ridges till the ridges are just one pixel widethis process is shown in figure 6[15].

centre of termination (ending) less than assumed distance, remove minutiae Process 2 : if distance between two centres of bifurcation is less than assumed distance, remove minutiae Process 3 : if distance between two centres of termination (ending) is less than distance, remove minutiae. Figure 9 shows the fingerprint image after the three processes.

Figure 6: Thinning Binary Image (Left) and Thinning image (Right)

E. Minutiae Marking:

Figure 9: after the three processes

After the fingerprint ridge thinning, marking minutiae points is relatively easy. By two steps: 1.Divide 1.Divide object into block. 2.Compute 2.Compute number of ones of 3x3 windows. Crossing number (Cn) is used to identify the minutiae points. It is defined as half of the sum of differences between intensity values of two adjacent adjace nt pixels. p ixels. If crossing number is 1, 2, 3 or greater, then the minutiae points are considered as ending, normal ridge, bifurcation respectively as shown in figure 7:

a) If central 1and has 1 one value neighbor, central pixel is termination (ending),Cn=1; b) If central 1 and has 3 one value neighbor, central pixel is bifurcation, Cn=2; c) If central is 1 and has 2 one value neighbor, central pixel is usual pixel, Cn=3.

4.2 Fingerprint Matching Algorithm (Minutiae-based Matching):

An effective fingerprint matching algorithm is proposed based on feature extraction. Given two sets of bifurcation minutiae of two fingerprint images, the minutiae match algorithm determines whether the two bifurcation minutiae sets are from the same finger or not. An alignment-based match algorithm partially includes two consecutive stages: one is alignment stage and the second is match stage [12]. 1. Alignment stage: Given two fingerprint images to be matched, choose any one minutia from each image; calculate the similarity of the two ridges associated with the two referenced minutiae points. If the similarity is larger than a threshold, transform each set of minutiae to a new coordination system whose origin is at the referenced point and whose x-axis is coincident with the direction of the referenced point. Let B1 & B2 be the two bifurcation minutiae sets given by: B1= {m1, m2, mM} where mi= (xi, yi, θi) B2= {m1', m2', mN'} where mi'= (xi', yi', θi')

Figure 7: Minutiae Marking

Also the average inter-ridge width D is estimated at this stage. The average inter-ridge width refers to the average distance between two neighbouring ridges. Figure 8 shows the bifurcation and ending minutiae marking.

Where xi and yi are the minutia coordinates, θi is direction discrepancy, M is the total numbers of minutiae points in set B1 and N is the total numbers of minutiae points in set B2. We choose one minutiae from each set to find the ridge correlation factor between them. The ridge associated with each minutia is represented as a series of x-coordinates (x1, x2,…, xn) of the points on the ridge. A point is sampled per ridge length L starting from the minutiae point, where the L is the average inter-ridge length, where n is set to 10 unless the total ridge length is less than 10*L.



So the similarity of correlating the two ridges is derived from equation 1: (1)

Where (xi...xn) and (Xi…Xn) are the set of x-coordinates for each of the 2 minutiae chosen, and m is minimal one of the n and N value. We test all the similarity values from 0.1 to 0.9 and if the similarity score is larger than the similarity test value, then go to step 2, otherwise continue to match the next pair of ridges. Let F (x, y,θ) be reference minutiae found from step 1(say from B1). For each fingerprint, translate and rotate all other minutiae (xi, yi, θi) with respect to the F according with the equation 2 formulas: (2)

The new coordinate system is originated at reference minutiae F and the new x-axis is coincident with the direction of minutiae F. No scaling effect is taken into account by assuming two fingerprints from the same finger have nearly the same size. So we transformed sets of minutiae as B1’ & B2’. 2.

Match stage: The elastic match algorithm is used to

count the matched minutiae pairs by assuming two minutiae having nearly the same position and direction. The matching algorithm for the aligned minutiae patterns needs to be elastic since the strict match is impossible due to the slight deformations and inexact quantization of minutiae. The final match scores given by 3: (3)

The score is 100*ratio and ranges from 0 to 100. We use the threshold value as 80%, where this value is proven to be the most appropriate threshold value by many researches. Figure 10 shows the diagram of Minutiae based fingerprint recognition.

Figure 10 .Diagram of Minutiae based fingerprint recognition

5. EXPERIMENT RESULTS: Based on database provided by FVC2002 DB2 (Fingerprint Verification Competition 2002), we conduct an experiment to decide the optimum number of templates, where the threshold percentage of matching is 80 %. 5.1 Multi-biometric systems :

Most biometric systems that rely on single biometric indicator often report unacceptable errors either due to noisy inputs or the limitations of the data set itself. Information fusion can be a powerful tool capable of increasing the reliability of biometric systems. The main idea is to combine individual outcomes so the reliability of the overall system improved when compared to that of individual systems. Multi biometric systems attempt to enhance authentication performance by combining multiple evidence of the same identity [18]. In this paper, we used multiple evidence of the same biometric trait to improve the matching performance. The most common performance metrics are the False Acceptance Rate FAR and the False Rejection Rate FRR. They are very much dependent on the biometric used and on the technical implementation of the biometric solution. Furthermore the FRR is strongly person dependent. Taking these into account; one person is insufficient to establish an overall FRR for a solution. Also FRR might increase due to environmental conditions or incorrect use, for example when using dirty fingers on a fingerprint reader [17]. First it is required to determine the number of image template that gives best recognition performance. Figure 11 shows the FAR and FRR curves for one and two templates in the minutiae based algorithm. From the curves we have zero FAR at similarity 0.5 for both cases. We can notice that, increasing the number of training sets is useful in FRR, where with one and two templates at zero FAR we have FRR as 20 and 12 respectively.



normalization techniques are considered in our experiments. [19] Min-max normalization is best suited for the case where the bounds (maximum and minimum values) of the scores are known. In this case, we can easily shift the minimum and maximum scores to 0 and 1, respectively. Given a set of matching scores {sk}, k = 1, 2...n, the normalized scores are given by 4: (4)

Figure 11. FAR and FRR curves for the original algorithm .

5.2 Matching Results :

Min-max normalization retains the original distribution of scores except for a scaling factor. Distance scores can be transformed into similarity scores by subtracting the minmax normalized score from 1 [20]. The match score distribution curves for figures 14 and 15 is estimated by using genuine and impostor match results then all individual matching results are combined to form one single final score by using the min-max normalization.

In the following experiments, we compare the original and enhanced algorithms for two templates. Figure 12 and 13 show the FAR and FRR results for our enhanced minutiae based approach and the original approach. With enhanced algorithm, FAR starts to be zero at less similarity value (0.4), than original algorithm (0.5).

Figure 14: Normalized Genuines

Figure 12. FAR curves for two templates .

Figure 15: Normalized Impostors

Figure 13 FRR curves for two templates.

5.3 Normalized Genuine and Impostor curves

The fusion in a biometric system can be fulfilled at four different levels, i.e. sensor level, feature level, score level, and decision level. The score level fusion of minutiae and extended fingerprint features are most widely used in fingerprint recognition. In parallel fusion approaches, different fingerprint features are matched separately and simultaneously, each defining a matcher. Score normalization is then applied to change the location and scale parameters of the match score distributions at the outputs of the individual matchers. The min-max (MMN)

From the previous normalization results, when we increase the similarity values, the genuine and impostor results are decrease. We notice that at similarity value of 0.7, there is no FAR and more than half of the tested fingerprint images are accepted. 5.4 Detection error normalization

trade-off

(DET)

after

Biometric performance is often represented by a Detection Error Trade-off (DET) curve. It plots FAR versus FRR for a variable threshold and measures the overall performance of a biometric system over all possible thresholds. In figure 16, DET curves for our experiment are shown.



Figure 16: Detection Error Trade-off (DET)

6. CONCLUSIONS

In this paper we used minutiae based fingerprint recognition and rely on the bifurcation minutiae points not only in the alignment matching stage sets but we used it also at the ending minutiae points in the match stage. We conducted our experiments using the database provided by FVC2002 DB2 and chose seven persons with 3 images each. Our algorithm is coded using us ing MATLAB. MATLAB. To increase the system performance, we used more than one template where it is proven that using two templates in the training set fulfil zero FAR with acceptable less FRR. We conclude that, using two templates images, our bifurcation minutiae based algorithm ensures better performance matching results over the minutiae algorithm. We have more accurate results with zero FAR at less similarity value.

REFERENCES [1]

D. Maltoni, D. Maio, A. K. Jain and S. Prabhakar, Handbook of Fingerprint Recognition, Springe, 2003. [2] (2004) The Forensic science project website. [Online]. Available : http://sites.bergen.org/forensic/fingerprint.htm [3] D.Kocharyan and H. Sarukhanyan,"Feature Extraction Techniques and Minutiae-Based Fingerprint Recognition Process,"International Journal of Multimedia Technology (IJMT), vol.1,pp.31-35, 2011. [4] Z. Shihai and L. Xiangjiang,"Fingerprint Identification and Its Applications in Information Security Fields ," in International Conference of Information Science and Management Engineering, 2010,p 97 – 99. [5] (2011) The Advanced Source Code website. [Online]. Available: http://www.advancedsourcecode.com/fingerprintdatabase.asp. [6] (2010)Biometrics Ideal Test website. [Online]. Available:http://biometrics.idealtest.org/findTotalDbByMode.d o?mode=Fingerprint. [7] C. Watson and C. Wilson,"NIST Special Database 4 Fingerprint Database," 1992. [8] (2006)The Fingerprint verification Competition website. [Online]. Available : http://bias.csr.unibo.it/fvc2006/databases.asp [9] (2003)The Fingerprint verification Competition website. [Online]. Available :http://bias.csr.unibo.it/fvc2004/databases.asp [10] (2001)The Fingerprint verification Competition website. [Online]. Available :http://bias.csr.unibo.it/fvc2002/databases.asp [11] (2000)The Fingerprint verification Competition website. [Online]. Available :http://bias.csr.unibo.it/fvc2000/databases.asp [12] S.Bana, Dr. D. Kaur, "Fingerprint recognition using image

segmentation," in international journal of advanced engineering science and technology, vol No. 5, pp. 012 – 023, 2011. [13] K. Ito, A. Morita, T. Aoki, T. Higuchi , H. Nakajima and K. obayashi, "A Fingerprint Recognition Algorithm Using PhaseBased Image Matching for Low-Quality Fingerprints," in IEEE International Conference on Image Processing ,Sept.2005. [14] F.A. Afsar, M. Arifand M. Hussain, "Fingerprint Identification and Verification System using Minutiae Matching," National Conference on Emerging Technologies, 2004. [15] K. Abbad ,A. Aarab and H. Tairi, "Fingerprint "Fingerprint Verification Based On Minutiae andMoments,", 5th International Symposium on I/V Communications and Mobile Network (ISVC),2010 [16] M. Sepasian, C. Mares, S. M. Azimi, and W. Balachandran, "Image Enhancement for Minutiae-Based Fingerprint Identification ,"37th IEEE App lied Imagery Pattern Recognition Workshop, Oct. 2008 [17] (2010) biometric-solution website [Online]. Available: http://www.biometricsolutions.com/index.php?story=performance_biometrics [18] H.Tilborg and S.Jajodia, Encyclopaedia of Cryptography and Security, Springe, 2011. [19] Q.Zhao, F.Liu , L.Zhang and D.Zhang, "Parallel versus Hierarchical Fusion of Extended Fingerprint Features," 20th International Conference on Pattern Recognition (ICPR),pp. 1132 – 1135, Aug. 2010 [20] A. Ross, K. Nandakumar and A. K. Jain, Handbook of Multibiometrics, Springer, 2006. Dr. Manal Abdullah is currently an Assistant Professor in Department of Computer Science (Faculty of Computing and Information Technology, King Abdulaziz University). She has published several researches in reputable international journals and conferences. She participated in numerous research projects in KAU. Mona Alkhozae completed BS in Computer Science from King Abdulaziz University KAU in 2006. She is currently a master student in Computer Science in King Abdulaziz University KAU and working as demonstrator in the deanship of information technology over five years. Mashaiel Alsowaiel completed BS in Computer Science from King Abdulaziz University KAU in 2006. She is currently a master student in Computer Science in King Abdulaziz University KAU and working as demonstrator in the deanship of information technology over five years. years.


Fingerprint Matching Approach Based on Bifurcation Minutiae

Recommend Documents