SlideShare a Scribd company logo
1 of 8
Download to read offline
IOSR Journal of Computer Engineering (IOSRJCE)
ISSN: 2278-0661 Volume 3, Issue 3 (July-Aug. 2012), PP 39-46
www.iosrjournals.org

            Feature Level Fusion Based Bimodal Biometric Using
                    Transformation Domine Techniques
 1
    Ramachandra A C, 2Abhilash S K, 2K B Raja, 3Venugopal K R, 4L M Patnaik
       1
        Department of Electronics and Communication, Alpha College of Engineering, Bangalore, 560077.
2
    Department of Electronics and Communication, University Visveswaraya College of Engineering, Bangalore,
           3
             Principal, University Visveswaraya College of Engineering, Bangalore, Karnataka, India,
                  4
                    Honorary Professor, Indian Institute of Science, Bangalore, Karnataka, India,

Abstract: Bimodal biometric used to authenticate a person is more accurate compared to single biometric trait.
In this paper we propose Feature Level Fusion based Bimodal Biometric using Transformation Domine
Techniques (FLFBBT). The algorithm uses two physiological traits viz., Fingerprint and Face to identify a
person. The Region of Interest (ROI) of fingerprint is obtained using preprocessing. The features of fingerprint
are extracted using Dual Tree Complex Wavelet Transforms (DTCWT) by computing absolute values of high
and low frequency components. The final features of fingerprint are computed by applying log on concatenated
absolute value of high and low frequency components. The face image is preprocessed by cropping only face
part and Discrete Wavelet Transforms (DWT) is applied. The approximation band coefficients are considered as
features of face. The fingerprint and face image features are concatenated to derive final feature vector of
bimodal biometric. The Euclidian Distance (ED) is used in matching section to compare test biometric in the
database, it is observed that the values of EER and TSR are better in the case of proposed algorithm compared
to individual transformation domain techniques.
Keywords: Biometrics, Face, Fingerprint, DWT and DTCWT

                                         I.         INTRODUCTION
          The authentication of a person using biometric is accurate and robust compared to traditional methods
of identification. The security can be easily broken if the traditional methods of personal authentication using
password and ID cards since they can be stolen or lost. In presence of sophisticated biometric identification
system any unusual such as robbery, entry of an unauthorized person etc ., can be avoided. The biometric
identification system is based on characteristics features of human body such as eyes, face, fingerprint, DNA
etc., which are known as Physiological Biometrics which are constant throughout the life time and the behavior
of person which depends on age and mood of a person such as signature, keystroke etc., known as Behavioral
Biometrics, which are not constant throughout the life. The biometric system is operated in two different modes
based on the application such as (i) Verification Mode: it is one to one comparison of test biometric with the
template stored in database, (ii) Identification Mode: it is one to many comparisons of test with the database, if
the comparison is nearer to the predefined threshold value then the person is identified.
          The general biometric system has data acquisition section, preprocessing section, feature extraction
section, matching section. The function of data acquisition section is to collect the number of sample of
biometrics in different conditions and store it as a database. In preprocessing section the images are normalized
by colour converting, cropping and resizing. The features are extracted from normalized image in feature
extraction section by using spatial domine, transformation domine or combination of both. The final results are
obtained from extracted features in matching section by using the distance formulas like Euclidian Distance
(ED), Hamming Distance, Chi-square, Linear Discriminant Analysis, Support Vector Machine, Neural
Networks etc., The biometric systems are very much essential in applications like Home Security, Airport
Checking, Voting machine, Entry to high security zone like parliament House, ATM, Laptop, public places like
shopping mall etc.,
          The multimodal biometrics systems address the problem of single biometric system by considering
multiple evidences produced by multiple biometric traits. Multimodal biometric systems also provide anti-
poofing measures by making it difficult for an intruder to spoof multiple biometric traits simultaneously. An
integration fusion scheme is required to fuse the information presented by the individual modalities.
Contribution: In this paper FLFBBT algorithm is proposed to identify a person. The fingerprint and face of each
person is considered and features are extracted using DTCWT and DWT respectively. The ED is used in
matching section to authenticate a person.
          Organization: This paper is organized into following sections. Section 2 is an overview of Literature
Survey. The proposed model is described in Section 3. Section 4 discusses the Algorithm; Performance analysis
of the system is presented in Section 5 and Conclusions are contained in Section 6.
                                           www.iosrjournals.org                                         39 | Page
Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques
                               II.       LETRAERETURE SURVEY
          Wang Wenchao and Sun Limm [1] discussed a Fingerprint identification algorithm that combines point
match and image match. This algorithm depends on singular point abstraction and wavelet transform coefficient.
First singular points are extracted for rough match and image calibration. Ma Yinping and Huang Yongxing [2]
introduced a kind of wavelet transform adaptive threshold of the Fingerprint image denoising method. The
approach is one of the Fingerprint image wavelet decomposition, and then based on Bayes framework, selecting
different optimal threshold, combined with soft threshold value method to Fingerprint image denoising, improve
the Fingerprint image Peak Signal-to-Noise Ratio. Atif Iqbal and Anoop Namboodiri [3] explore the
effectiveness of weak features in a cascade for filtering fingerprint databases considering a set of potential
indexing features computed from minutiae triplets and minutiae quadruplets. Each stage of filtering consists of
projecting the probe onto a specific line and the removal of database samples outside a window around the
probe. Hao et al., [4] proposed an algorithm based on multiple minutiae partitions to match general distorted
fingerprints. Minutiae in an input and a template fingerprint are globally partitioned into pairs of multiple
neighbors, each of which consist of similar local minutia topology and features. The minutia similarity is
additionally evaluated by investigating ridge shape and the geometric similarities of the corresponding neighbor
minutiae. Jinhai Zhang [5] discussed methods of fingerprint image segmentation. Some of the segmentation
methods are classified based on block variance method, Pattern methods, and synthesis method. The grading
segmentation method automatically identifies some close value in a hierarchical segmentation.
          Arun et al., [6] proposed the feature extraction techniques for three modalities viz. Fingerprint, iris and
Face. The extracted information from each modality is stored as a template. The information are fused at the
match score level using a density based score level fusion, Gaussian Mixture Model (GMM) followed by the
Likelihood ratio test. Ngoc-Son Vu and Alice Caplier [7] invented a feature descriptor set called Patterns of
Oriented Edge Magnitudes (POEM) for feature extraction by applying the self-similarity operator on
accumulated edge magnitudes across different directions. Based on these Features, two feature descriptors are
constructed. Carlos et al., [8] proposed an Active Shape Models (ASM) landmark selection scheme to improve
the ASM performance in face recognition applications. The proposed scheme selects robust landmark points
where relevant facial features are found and assigns higher weights to their corresponding features in the face
classification stage. Xiaoni Liang and Weiqing Tong [9] presented an accurate and fast approach to estimate the
pose of face in near-infrared images. The approach automatically extracts the feature points from a normalized
face image. The pupils and nose tip are used to estimate the pose rotation of yaw and the pupils, nose tip and the
center line of mouth is used to estimate the pose rotation of pitch.
          Ajay Kumar and Yingbo Zhou [10] proposed a system which simultaneously acquires the finger-vein
and low-resolution fingerprint images and combines these two evidences using a score-level combination
strategy. The two score-level combinations are holistic and nonlinear fusion. Kolhandai et al., [11] present an
intelligent hybrid features based face recognition method which combines the local and global approaches to
produce a complete a robust and high success rate face recognition system. The global features are computed
using principal component analysis while the local features are ascertained configuring the central moment,
Eigen vectors and the standard deviation of the eyes, nose and mouth segments of the human face as the
decision support entities of the Generalized Feed Forward Artificial Neural Network. Ramya et al., [12]
proposed a system with key is generated from the fused feature which is of greater than 128 bit which is enough
for AES encryption, Hash Encoding and AES decryption is done. Message is subjected to hash and hash of
message is obtained and it is given to AES encryption as a message and the key is obtained from the fused
feature.
          Lili Liu and Tianjie Cao [13] defined an efficient verification system based on biometrics, which
access real system resources without physical key. Based on the pattern matching, the feature are extracted
from the fingerprint images, the feature depends on the position and angle of the input fingerprints. The ridges
and valleys structure information integration is employed to perform the fingerprint minutiae alignment. Biggio
et at., [14] have investigated the robustness of different score fusion rules for multimodal biometric verification
systems, against spoofing attacks for Fingerprint and Face biometrics. The simulated worst-case scenario
considered is not representative of the score distribution of real spoofing attacks. Abhishek Nagar et at., [15]
discussed a feature-level fusion framework for the design of multi biometric cryptosystems that simultaneously
protects the multiple templates of a user using a single secure sketch. The feasibility of such a framework has
been demonstrated using both fuzzy vault and fuzzy commitment.

                                             III.         MODEL
         In this section the block diagram of the proposed model as shown in Figure 1 is discussed. The two
biometric traits viz., Fingerprint and Face of a person are considered together to identify a person. The transform
domain techniques such as DTCWT and DWT are used to extract Features from Fingerprint and Face
respectively.

                                                www.iosrjournals.org                                       40 | Page
Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques
                       Database Section                                 Test
                                                                     Section
                 Fingerprint               Face             Fingerprint              Face


                 Preprocess             Preprocess         Preprocess             Preprocess

                  DTCWT                   DWT                DTCWT                  DWT


                  Features               Features            Features              Features


                               Fusion                                    Fusion

                                               Match/Mismatch


                                              Accept/Reje
                                              ct
                                 Figure 1: Block diagram of proposed algorithm

3.1. DATABASE:
3.1.1. Fingerprint database: The DB3 Fingerprint database from FVC2004 competitions is considered to test the
algorithm. The database consists of 110 persons with 8 Fingerprint images per person having image size of
300x480 with a resolution of 512 dpi. The Fingerprint images are collected in three different sessions with at
least two weeks’ time separating each session using thermal sweeping sensor [16]. The DB3 has eight images of
single person shown in Figure 2. The Fingerprint database is created in our proposed model by considering
minimum of 10 persons and maximum of 40 persons with 6 images per persons. The number of images with
database varies between 60 and 240. The Seventh and Eighth Fingerprint images of each person used in the test
section. The number of persons considered with out off database varies from 5 to 30 with one Fingerprint
images per person.




                                 Figure 2: Fingerprint Images of DB3 FVC2004

 3.1.2. Face database: The Olivetti Research Laboratory (ORL) gray scale [17] is considered to test the
proposed model. The data base has 40 persons and 10 images per person i.e., 400 Face images in the database.
The each Face image has frontal view with slight tilt of head with a size of 92 x 112. To test the proposed
modal, the Face data base is created by considering 10 persons to 40 persons with six images per person. The
seventh and eight images of each person are considered as test images. The out off database has five persons to
30 persons with one Face image per person is used to compute FAR. The Face image samples of one person are
shown in figure 3.




                                           Figure 3: Face Images of ORL

                                                  www.iosrjournals.org                                41 | Page
Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques
1.2.      PREPROCESSING:
         The Fingerprint image of size 300 x 480 is manually cropped to a dimension of 201 x 401 to obtain
Region of Interest (ROI) Fingerprint area by eliminating four side edges. The size of ROI is resized to 256 x 256
to apply DTCWT for Features. The original Fingerprint image and its cropped and resized image are shown in
figure 4. The original Face image of size 92 x 112 is manually cropped to obtain ROI of size 71 x 93 by
eliminating edge of Face image. The ROI is resized to an arbitrary value say size of 192 x 192 to apply DWT to
derive Features. The original Face image and its cropped and resized image are shown in figure 5.




       (a)               (b)                                          (a)                                     (b)
Figure 4: (a) Original Fingerprint Image (300 x 480)                Figure 5: (a) Original Face Image (92 x 112)
          (b) Preprocessed Fingerprint Image (256 x 256)                     (b) Preprocessed Face Image (192 x 192)

1.3.      FEATURE EXTRACTION:
          The Transform Domain Features from the Fingerprint and Face image are obtained by using the Dual
Tree Complex Wavelet Transforms (DTCWT) and Discrete Wavelet Transforms (DWT) transforms
respectively. The logarithm is applied on DTCWT magnitudes of Fingerprint to obtain Final Features of
Fingerprint image. The Final Fingerprint Features and Features of Face are fused by concatenating to derive the
Final set of Features to test the algorithm.

3.3.1. DTCWT: It produces complex coefficients by using dual tree wavelet filters and gives real and imaginary
parts, which consists of both low frequency and high frequency bands. The high frequency band is ignored and
low frequency band is considered and decomposed into two parts for next levels and process is continued until
the required six levels as shown in Figure 6. At level six all the low frequency coefficients and high frequency
coefficients from both real and imaginary parts are considered. The magnitudes of both real and imaginary parts
are calculated using Equations 1 and 2 respectively and have a dimension of 1 x 48. Magnitudes are
concatenated as given in Equation 3 to obtain a single Vector with a dimension of 1 x 96. For the obtained
single Vector logarithmic scale is applied as given in Equation 4 to normalize the magnitude, which is
considered as Feature Vector FFP for Fingerprint image.

M L  [( RL 6) 2  ( I L 6) 2 ]                                         …………………. (1)

                                  M H  [( RH 6) 2  ( I H 6) 2 ]                       …….……………. (2)

                              M  [M H : M L ]                                        …….…………….. (3)

                             FFP  Log(M )                                              ……………………. (4)

3.3.2 DWT: The preprocessed Face image of size 192 x 192 is considered and one level haar wavelet is applied
using base Equation 5 to derive four sub bands such as Approximation band, Horizontal band, Vertical band and
Diagonal band as shown in Figure 7. The Approximation band has low frequency components and has
significant information of Face image. The Horizontal, Vertical and Diagonal bands are high frequency bands
and have in significant information of Face image like edge information’s. The approximation band coefficients
are considered as Features of Face image FF which is of dimension 1 x 9216.




                                            Figure 6: DWT sub bands of Face image.
                                                        www.iosrjournals.org                                42 | Page
Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques
          1 0  t 1/ 2
         
 (t )    1 1/ 2  t 1                                                  ………………………… (5)
          0 otherwise
         

1.4.      FUSION OF FEATURES:
          The Log DTCWT feature of Fingerprint FFP and Approximation band coefficient FF of DWT of Face
image are concatenated to obtain Final feature Vector set for bimodal biometrics. The Final feature Vector FFV
is given in Equation six Equation 5. The Final feature Vector size is of 1x 9312.

                                                                                                        Level 6
Level 5
                                                                                               RL6
                      Real Part                                            Level 4       RL5
       Level 3                                                            RL4
        Level 2                                             RL3                                RH
                  Level 1                                                                RH    6
                                            RL2
                   RL1                                                    RH             5
                                                            RH            4
                                            RH              3                                  IL6
                   RH                       2                                            IL5
                   1                                                      IL4
                                                            IL3                                IH6
       X                                                                                 IH5
                                            IL2
                                                                           IH4
                     IL1                                    IH3
                                            IH2
                     IH1
                                                                                                Imaginary Part
                                                  Figure 6: Different levels of DTCWT.

FFV  [ FFP : FF ]                                                ………………... (6)

1.5.     MATCHING:
         The Final fused feature Vector derived from both Fingerprint and Face images are considered to test
match/non match of a person. The Final fused Features of test image are compared with Final fused Features of
database images using Euclidian distance as given in Equation 7.

                  
              1      M
d ( p, q)                  ( pi  qi ) 2                                   ……………. (7)
              M      i 1

M = Dimension of Vector.
Pi = Data Feature Vector
Qi = Test Feature Vector
The TSR, FRR and FAR values are computed by compared ED values with predefined thresh hold values.

                                          IV.          ALGORITHM
          Problem Definition: The proposed bimodal biometric algorithm using Log (DTCWT) and DWT to
identify a person is discussed and is given in Table 1
The Objectives:
 The bimodal biometric recognition is identified using fusion of DTCWT and DWT coefficients.
 To reduce FAR,FRR and EER
 To increase TSR




                                                           www.iosrjournals.org                        43 | Page
Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques
                                     Table 1: Algorithm of proposed System.
     INPUT: Fingerprint and Face images.
     OUTPUT: Matched /Non-matched Test image

     1. Resize Fingerprint image to 256x256.
     2. DTCWT is applied on Fingerprint of database and test Fingerprint.
     3. The Fingerprint initial features are obtained by concatenation of magnitudes of low and high
        frequency components.
     4. The Fingerprint final features are derived by Log of initial features.
     5. Resizing of Face image to 192x192.
     6. DWT is applied on database images and test image.
     7. The Face features are obtained by considering coefficients of approximate bands.
     8. The bimodal biometric final features are derived by concatenation of Fingerprint and Face features.
     9. The final features of test image are compared with final features of images in the database using ED.

                                   V.          PERFORMANCE ANALYSIS:
         The Fingerprint databaseFVC2004 DB3 and Face database ORL are considered to test the proposed
algorithm in terms of the following performance parameters.

1.     False Acceptance Rate (FAR):
       It is the probability that an unauthorized person is incorrectly accepted as authorized person and it is
computed using Equation 8
                   FAR = Number of unauthorized persons Identified
                        Total Number of unauthorized persons                  ………………………… (8)

2.         False Rejection Rate (FRR):
It is the probability that the authorized person is incorrectly rejected and it is computed using the Equation 9.
FRR = Number of authorized persons rejected
Total Number of authorized persons                          ……………………….. (9)

3.        Equal Error Rate (EER):
 It is the rate at which both FAR and FRR are equal.

4.       Total Success Rate (TSR):
         It is the rate at which match occurs successfully. The Numbers of persons recognized correctly in the
database is given in the Equation 10.
%TSR = Number of persons recognized correctly * 100
                            Total Number of persons in the database            …………………………. (10)
         The proposed algorithm is robust in recognition of a person as it uses two biometric traits. The
variation of FRR, FAR and percentage TSR with threshold is tabulated in the Table 2 As threshold value
increases the FRR values decreases from 1 to zero whereas FAR and TSR values increases .The maximum
percentage TSR value of 90% is achieved from 0.7 threshold value.
         The variation of EER and TSR values for test images 7 and 8 of each person with Persons Inside
Database (PIDB): Persons Outside Database (PODB) are given in Table 3.The values of EER increases with
PODB and PIDB.The values of EER and TSR depends on the test images, i.e., for test image number 7, EER is
0.16 and TSR is 70% for PIDB and PODB combination of 10:30 and for test image number 8, EER is 0.13 and
TSR is 90%. For PIDB and PODB combination10:30.

                             Table 2: FRR, FAR and % TSR for different Threshold.
                                     Thresh Hold   FRR FAR %TSR
                                     0             1      0        0
                                     0.05          1      0        0
                                     0.25          1      0        0
                                     0.3           0.9    0        10
                                     0.4           0.8    0        20
                                     0.45          0.6    0        40
                                     0.5           0.5    0        50

                                                www.iosrjournals.org                                       44 | Page
Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques
                                   0.55               0.4      0.033   60
                                   0.6                0.2      0.033   70
                                   0.65               0.2      0.1     70
                                   0.7                0        0.2     90
                                   0.75               0        0.4     90
                                   0.8                0        0.633   90
                                   0.85               0        0.733   90
                                   0.9                0        0.833   90
                                   0.95               0        0.9     90
                                   1                  0        0.96    90
               Table 3: Comparison of EER and Percentage of TSR for different PIDB: PODB.


         The variations of FAR and FRR with threshold for PIDB and PODB combination 10:30 is given in
Figure 7. As threshold increases FRR decreases and FAR increases the EER value of 0.13 is intersection of FRR
and FAR plot.




                          Figure 7: Plot of FAR and FRR V/S Thresh Hold with Log

         The values of EER and TSR for DTCWT, DWT and combinations of LOG (DTCWT) and DWT i e.,
proposed algorithm is given in Table 4. It is observed that the proposed algorithm has low EER value compared
to individual DTCWT and DWT algorithms. The value of percentage TSR is high in the case of proposed
algorithm compared to individual transformation algorithms. Hence the proposed bimodal biometric algorithm
has better efficiency and also robust as two biometric traits are used to identify a person compared to single
biometric trait.

    Table 4: Comparison of EER and Percentage of TSR for different Transforms and Proposed Algorithm.
                                                             Method
                           Parameters         DTCWT          DWT            Proposed
                                                                            FLFBBT
                           EER            0.18                0.485    0.13
                           %TSR           80                40         90


                                        VI.         CONCLUSION:
         In this paper FLFBBT algorithm is proposed to identify a person accurately. The DTCWT is applied on
ROI of fingerprint and computed magnitudes of low and high frequency components. The features of fingerprint
are derived by applying log on concatenated magnitudes of low and high frequency components. The DWT is
applied on face image and features are extracted by coefficients of approximate band. The final features of
bimodal biometrics are computed by concatenating fingerprint and face features. The ED is used for comparison
to authenticate a person. It is observed that the values of EER and TSR are better in the proposed algorithm
compared to individual transform domain techniques. In future the algorithm can be verified by using both
spatial domain techniques and transform domain techniques at different levels of fusion.




                                               www.iosrjournals.org                                  45 | Page
Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques
                                                         REFERENCES
[1]    Wang Wenchao and Sun Limin, “A Fingerprint Identification Algorithm Based on Wavelet Transformation Characteristic
       Coefficient,” International Conference on Systems and Informatics, pp1-3, 2012.
[2]    Ma Yinping and Huang Yongxing, “Adaptive Threshold Based on Wavelet Transform Fingerprint Image Denoising,” International
       Conference on Computer Science and Electronics Engineering, pp. 494- 497, 2012.
[3]    Atif Iqbal and Anoop Namboodiri, “Cascaded Filtering for Fingerprint Identification using Random Projections,” Computer Society
       Conference on Computer Vision and Pattern Recognition Workshops, pp. 77- 82, 2012.
[4]    Hao NI, Dongju LI, Tsuyoshi Isshiki and Hiroaki Kunieda, “Robust Multiple Minutiae Partitions for Fingerprint Authentication,”
       International Symposium on Biometrics and Security Technologies, pp. 35- 44, 2012.
[5]    Jinhai Zhang, “The research of fingerprint image Segmentation method,” Second International Conference on Consumer
       Electronics, Communications and Networks, pp. 701- 704, 2012.
[6]    S ArunVivek, J Aravinth and S Valarmathy, “Feature Extraction for Multimodal Biometric and Study of Fusion Using Gaussian
       Mixture Model,”Proceedings of the International Conference on Pattern Recognition, Informatics and Medical Engineering, pp.
       391-392, 2012.
[7]    Ngoc-Son Vu and Alice Caplier, “Enhanced Patterns of Oriented Edge Magnitudes for Face Recognition and Image Matching”
       IEEE Transactions on Image Processing, vol. 21, no. 3, 2012.
[8]    Carlos A. R. Behaine and Jacob Scharcanski, “Enhancing the Performance of Active Shape Models in Face Recognition
       Applications,” IEEE Transactions on Instrumentation and Measurement, vol. 61, no. 8, 2012.
[9]    Xiaoni Liang and Weiqing Tong, “Face Pose Estimation Using Near-Infrared Images,” International Conference on
       Communication Systems and Network Technologies, pp. 216- 220, 2012.
[10]   Ajay Kumar and Yingbo Zhou, “Human Identification Using Finger Images,” Transactions on Image Processing, vol. 21, no. 4,
       April 2012.
[11]   Kolhandai Yesu, Himadri Jyoti Chakravorty, Prantik Bhuyan, Rifat Hussain and Kaustubh Bhattacharyya, “Hybrid Features Based
       Face Recognition Method Using Artificial Neural Network,” Second National Conference on Computational Intelligence and Signal
       Processing, pp. 40 – 46, 2012.
[12]   M.Ramya, A. Muthukumar and S.Kannan, “Multi biometric based Authentication using Feature Level Fusion,” International
       Conference on Advances In Engineering, Science And Management, pp. 191- 195, 2012.
[13]   Lili Liu and Tianjie Cao, “The Research and Design of an Efficient Verification System Based on Biometrics,” International
       Conference on Computer Science and Electronics Engineering, pp. 707-710, 2012.
[14]   B Biggio, Z Akhtar, G Fumera, G L Marcialis and F Roli, “Security Evaluation of Biometric Authentication Systems Under Real
       Spoofing attacks,” IET Biometrics, Vol. 1, pp. 11–24, 2012.
[15]   Abhishek Nagar, Karthik Nandakumar and Anil K Jain, “Multibiometric Cryptosystems Based on Feature Level Fusion,” IEEE
       Transactions on Information Forensics and Security, vol. 7, no.1, pp. 225-268, 2012.
[16]   http://bias.csr.unibo.it/fvc2004/default.asp, FVC2004-Third International Fingerprint Verification Competition.
[17]   Ferdinando Samaria and Andy Harter, “Parameterization of a Stochastic Model for Human Face Identification,” Proceedings of
       Second IEEE Workshop on Applications of Computer Vision, December 1994.




                                                       www.iosrjournals.org                                                46 | Page

More Related Content

What's hot

International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)inventionjournals
 
Transform Domain Based Iris Recognition using EMD and FFT
Transform Domain Based Iris Recognition using EMD and FFTTransform Domain Based Iris Recognition using EMD and FFT
Transform Domain Based Iris Recognition using EMD and FFTIOSRJVSP
 
MultiModal Identification System in Monozygotic Twins
MultiModal Identification System in Monozygotic TwinsMultiModal Identification System in Monozygotic Twins
MultiModal Identification System in Monozygotic TwinsCSCJournals
 
A Hybrid Approach to Face Detection And Feature Extraction
A Hybrid Approach to Face Detection And Feature ExtractionA Hybrid Approach to Face Detection And Feature Extraction
A Hybrid Approach to Face Detection And Feature Extractioniosrjce
 
11.0002www.iiste.org call for paper.human verification using multiple fingerp...
11.0002www.iiste.org call for paper.human verification using multiple fingerp...11.0002www.iiste.org call for paper.human verification using multiple fingerp...
11.0002www.iiste.org call for paper.human verification using multiple fingerp...Alexander Decker
 
2.human verification using multiple fingerprint texture 7 20
2.human verification using multiple fingerprint texture 7 202.human verification using multiple fingerprint texture 7 20
2.human verification using multiple fingerprint texture 7 20Alexander Decker
 
Handwritten Character Recognition: A Comprehensive Review on Geometrical Anal...
Handwritten Character Recognition: A Comprehensive Review on Geometrical Anal...Handwritten Character Recognition: A Comprehensive Review on Geometrical Anal...
Handwritten Character Recognition: A Comprehensive Review on Geometrical Anal...iosrjce
 
An interactive image segmentation using multiple user input’s
An interactive image segmentation using multiple user input’sAn interactive image segmentation using multiple user input’s
An interactive image segmentation using multiple user input’seSAT Publishing House
 
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...CSCJournals
 
Bimodal Biometric System using Multiple Transformation Features of Fingerprin...
Bimodal Biometric System using Multiple Transformation Features of Fingerprin...Bimodal Biometric System using Multiple Transformation Features of Fingerprin...
Bimodal Biometric System using Multiple Transformation Features of Fingerprin...IDES Editor
 
Highly Secured Bio-Metric Authentication Model with Palm Print Identification
Highly Secured Bio-Metric Authentication Model with Palm Print IdentificationHighly Secured Bio-Metric Authentication Model with Palm Print Identification
Highly Secured Bio-Metric Authentication Model with Palm Print IdentificationIJERA Editor
 
IRJET- Facial Expression Recognition
IRJET- Facial Expression RecognitionIRJET- Facial Expression Recognition
IRJET- Facial Expression RecognitionIRJET Journal
 
An Indexing Technique Based on Feature Level Fusion of Fingerprint Features
An Indexing Technique Based on Feature Level Fusion of Fingerprint FeaturesAn Indexing Technique Based on Feature Level Fusion of Fingerprint Features
An Indexing Technique Based on Feature Level Fusion of Fingerprint FeaturesIDES Editor
 
Iris Encryption using (2, 2) Visual cryptography & Average Orientation Circul...
Iris Encryption using (2, 2) Visual cryptography & Average Orientation Circul...Iris Encryption using (2, 2) Visual cryptography & Average Orientation Circul...
Iris Encryption using (2, 2) Visual cryptography & Average Orientation Circul...AM Publications
 
Reduction of False Acceptance Rate Using Cross Validation for Fingerprint Rec...
Reduction of False Acceptance Rate Using Cross Validation for Fingerprint Rec...Reduction of False Acceptance Rate Using Cross Validation for Fingerprint Rec...
Reduction of False Acceptance Rate Using Cross Validation for Fingerprint Rec...IJTET Journal
 
International Journal of Computational Engineering Research(IJCER)
International Journal of Computational Engineering Research(IJCER)International Journal of Computational Engineering Research(IJCER)
International Journal of Computational Engineering Research(IJCER)ijceronline
 

What's hot (19)

An optimal face recoginition tool
An optimal face recoginition toolAn optimal face recoginition tool
An optimal face recoginition tool
 
International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)
 
Transform Domain Based Iris Recognition using EMD and FFT
Transform Domain Based Iris Recognition using EMD and FFTTransform Domain Based Iris Recognition using EMD and FFT
Transform Domain Based Iris Recognition using EMD and FFT
 
MultiModal Identification System in Monozygotic Twins
MultiModal Identification System in Monozygotic TwinsMultiModal Identification System in Monozygotic Twins
MultiModal Identification System in Monozygotic Twins
 
A Hybrid Approach to Face Detection And Feature Extraction
A Hybrid Approach to Face Detection And Feature ExtractionA Hybrid Approach to Face Detection And Feature Extraction
A Hybrid Approach to Face Detection And Feature Extraction
 
11.0002www.iiste.org call for paper.human verification using multiple fingerp...
11.0002www.iiste.org call for paper.human verification using multiple fingerp...11.0002www.iiste.org call for paper.human verification using multiple fingerp...
11.0002www.iiste.org call for paper.human verification using multiple fingerp...
 
2.human verification using multiple fingerprint texture 7 20
2.human verification using multiple fingerprint texture 7 202.human verification using multiple fingerprint texture 7 20
2.human verification using multiple fingerprint texture 7 20
 
Handwritten Character Recognition: A Comprehensive Review on Geometrical Anal...
Handwritten Character Recognition: A Comprehensive Review on Geometrical Anal...Handwritten Character Recognition: A Comprehensive Review on Geometrical Anal...
Handwritten Character Recognition: A Comprehensive Review on Geometrical Anal...
 
An interactive image segmentation using multiple user input’s
An interactive image segmentation using multiple user input’sAn interactive image segmentation using multiple user input’s
An interactive image segmentation using multiple user input’s
 
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...
 
Bimodal Biometric System using Multiple Transformation Features of Fingerprin...
Bimodal Biometric System using Multiple Transformation Features of Fingerprin...Bimodal Biometric System using Multiple Transformation Features of Fingerprin...
Bimodal Biometric System using Multiple Transformation Features of Fingerprin...
 
N010226872
N010226872N010226872
N010226872
 
A04820104
A04820104A04820104
A04820104
 
Highly Secured Bio-Metric Authentication Model with Palm Print Identification
Highly Secured Bio-Metric Authentication Model with Palm Print IdentificationHighly Secured Bio-Metric Authentication Model with Palm Print Identification
Highly Secured Bio-Metric Authentication Model with Palm Print Identification
 
IRJET- Facial Expression Recognition
IRJET- Facial Expression RecognitionIRJET- Facial Expression Recognition
IRJET- Facial Expression Recognition
 
An Indexing Technique Based on Feature Level Fusion of Fingerprint Features
An Indexing Technique Based on Feature Level Fusion of Fingerprint FeaturesAn Indexing Technique Based on Feature Level Fusion of Fingerprint Features
An Indexing Technique Based on Feature Level Fusion of Fingerprint Features
 
Iris Encryption using (2, 2) Visual cryptography & Average Orientation Circul...
Iris Encryption using (2, 2) Visual cryptography & Average Orientation Circul...Iris Encryption using (2, 2) Visual cryptography & Average Orientation Circul...
Iris Encryption using (2, 2) Visual cryptography & Average Orientation Circul...
 
Reduction of False Acceptance Rate Using Cross Validation for Fingerprint Rec...
Reduction of False Acceptance Rate Using Cross Validation for Fingerprint Rec...Reduction of False Acceptance Rate Using Cross Validation for Fingerprint Rec...
Reduction of False Acceptance Rate Using Cross Validation for Fingerprint Rec...
 
International Journal of Computational Engineering Research(IJCER)
International Journal of Computational Engineering Research(IJCER)International Journal of Computational Engineering Research(IJCER)
International Journal of Computational Engineering Research(IJCER)
 

Viewers also liked (9)

A0350107
A0350107A0350107
A0350107
 
G0314043
G0314043G0314043
G0314043
 
C0121116
C0121116C0121116
C0121116
 
B0311219
B0311219B0311219
B0311219
 
C0351725
C0351725C0351725
C0351725
 
B0340710
B0340710B0340710
B0340710
 
B0110508
B0110508B0110508
B0110508
 
D0362035
D0362035D0362035
D0362035
 
F0112937
F0112937F0112937
F0112937
 

Similar to G0333946

Feature Level Fusion Based Bimodal Biometric Using Transformation Domine Tec...
Feature Level Fusion Based Bimodal Biometric Using  Transformation Domine Tec...Feature Level Fusion Based Bimodal Biometric Using  Transformation Domine Tec...
Feature Level Fusion Based Bimodal Biometric Using Transformation Domine Tec...IOSR Journals
 
IJERD (www.ijerd.com) International Journal of Engineering Research and Devel...
IJERD (www.ijerd.com) International Journal of Engineering Research and Devel...IJERD (www.ijerd.com) International Journal of Engineering Research and Devel...
IJERD (www.ijerd.com) International Journal of Engineering Research and Devel...IJERD Editor
 
Sum Rule Based Matching Score Level Fusion of Fingerprint and Iris Images for...
Sum Rule Based Matching Score Level Fusion of Fingerprint and Iris Images for...Sum Rule Based Matching Score Level Fusion of Fingerprint and Iris Images for...
Sum Rule Based Matching Score Level Fusion of Fingerprint and Iris Images for...IRJET Journal
 
An overview of face liveness detection
An overview of face liveness detectionAn overview of face liveness detection
An overview of face liveness detectionijitjournal
 
Performance Enhancement Of Multimodal Biometrics Using Cryptosystem
Performance Enhancement Of Multimodal Biometrics Using CryptosystemPerformance Enhancement Of Multimodal Biometrics Using Cryptosystem
Performance Enhancement Of Multimodal Biometrics Using CryptosystemIJERA Editor
 
Iris recognition based on 2D Gabor filter
Iris recognition based on 2D Gabor filterIris recognition based on 2D Gabor filter
Iris recognition based on 2D Gabor filterIJECEIAES
 
Paper id 25201496
Paper id 25201496Paper id 25201496
Paper id 25201496IJRAT
 
A survey paper on various biometric security system methods
A survey paper on various biometric security system methodsA survey paper on various biometric security system methods
A survey paper on various biometric security system methodsIRJET Journal
 
Paper id 25201441
Paper id 25201441Paper id 25201441
Paper id 25201441IJRAT
 
An Improved Self Organizing Feature Map Classifier for Multimodal Biometric R...
An Improved Self Organizing Feature Map Classifier for Multimodal Biometric R...An Improved Self Organizing Feature Map Classifier for Multimodal Biometric R...
An Improved Self Organizing Feature Map Classifier for Multimodal Biometric R...ijtsrd
 
MULTI SCALE ICA BASED IRIS RECOGNITION USING BSIF AND HOG
MULTI SCALE ICA BASED IRIS RECOGNITION USING BSIF AND HOG MULTI SCALE ICA BASED IRIS RECOGNITION USING BSIF AND HOG
MULTI SCALE ICA BASED IRIS RECOGNITION USING BSIF AND HOG sipij
 
A NOVEL BINNING AND INDEXING APPROACH USING HAND GEOMETRY AND PALM PRINT TO E...
A NOVEL BINNING AND INDEXING APPROACH USING HAND GEOMETRY AND PALM PRINT TO E...A NOVEL BINNING AND INDEXING APPROACH USING HAND GEOMETRY AND PALM PRINT TO E...
A NOVEL BINNING AND INDEXING APPROACH USING HAND GEOMETRY AND PALM PRINT TO E...ijcsa
 
IRDO: Iris Recognition by fusion of DTCWT and OLBP
IRDO: Iris Recognition by fusion of DTCWT and OLBPIRDO: Iris Recognition by fusion of DTCWT and OLBP
IRDO: Iris Recognition by fusion of DTCWT and OLBPIJERA Editor
 
Biometric key generation using crow algorithm
Biometric key generation using crow algorithmBiometric key generation using crow algorithm
Biometric key generation using crow algorithmnooriasukmaningtyas
 
Palmprint and Handgeometry Recognition using FAST features and Region properties
Palmprint and Handgeometry Recognition using FAST features and Region propertiesPalmprint and Handgeometry Recognition using FAST features and Region properties
Palmprint and Handgeometry Recognition using FAST features and Region propertiesEditor IJCATR
 
Face Liveness Detection for Biometric Antispoofing Applications using Color T...
Face Liveness Detection for Biometric Antispoofing Applications using Color T...Face Liveness Detection for Biometric Antispoofing Applications using Color T...
Face Liveness Detection for Biometric Antispoofing Applications using Color T...rahulmonikasharma
 
International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)IJERD Editor
 

Similar to G0333946 (20)

Feature Level Fusion Based Bimodal Biometric Using Transformation Domine Tec...
Feature Level Fusion Based Bimodal Biometric Using  Transformation Domine Tec...Feature Level Fusion Based Bimodal Biometric Using  Transformation Domine Tec...
Feature Level Fusion Based Bimodal Biometric Using Transformation Domine Tec...
 
IJERD (www.ijerd.com) International Journal of Engineering Research and Devel...
IJERD (www.ijerd.com) International Journal of Engineering Research and Devel...IJERD (www.ijerd.com) International Journal of Engineering Research and Devel...
IJERD (www.ijerd.com) International Journal of Engineering Research and Devel...
 
Sum Rule Based Matching Score Level Fusion of Fingerprint and Iris Images for...
Sum Rule Based Matching Score Level Fusion of Fingerprint and Iris Images for...Sum Rule Based Matching Score Level Fusion of Fingerprint and Iris Images for...
Sum Rule Based Matching Score Level Fusion of Fingerprint and Iris Images for...
 
An overview of face liveness detection
An overview of face liveness detectionAn overview of face liveness detection
An overview of face liveness detection
 
D56021216
D56021216D56021216
D56021216
 
Performance Enhancement Of Multimodal Biometrics Using Cryptosystem
Performance Enhancement Of Multimodal Biometrics Using CryptosystemPerformance Enhancement Of Multimodal Biometrics Using Cryptosystem
Performance Enhancement Of Multimodal Biometrics Using Cryptosystem
 
Iris recognition based on 2D Gabor filter
Iris recognition based on 2D Gabor filterIris recognition based on 2D Gabor filter
Iris recognition based on 2D Gabor filter
 
Ijebea14 276
Ijebea14 276Ijebea14 276
Ijebea14 276
 
Paper id 25201496
Paper id 25201496Paper id 25201496
Paper id 25201496
 
A survey paper on various biometric security system methods
A survey paper on various biometric security system methodsA survey paper on various biometric security system methods
A survey paper on various biometric security system methods
 
Paper id 25201441
Paper id 25201441Paper id 25201441
Paper id 25201441
 
An Improved Self Organizing Feature Map Classifier for Multimodal Biometric R...
An Improved Self Organizing Feature Map Classifier for Multimodal Biometric R...An Improved Self Organizing Feature Map Classifier for Multimodal Biometric R...
An Improved Self Organizing Feature Map Classifier for Multimodal Biometric R...
 
MULTI SCALE ICA BASED IRIS RECOGNITION USING BSIF AND HOG
MULTI SCALE ICA BASED IRIS RECOGNITION USING BSIF AND HOG MULTI SCALE ICA BASED IRIS RECOGNITION USING BSIF AND HOG
MULTI SCALE ICA BASED IRIS RECOGNITION USING BSIF AND HOG
 
G041041047
G041041047G041041047
G041041047
 
A NOVEL BINNING AND INDEXING APPROACH USING HAND GEOMETRY AND PALM PRINT TO E...
A NOVEL BINNING AND INDEXING APPROACH USING HAND GEOMETRY AND PALM PRINT TO E...A NOVEL BINNING AND INDEXING APPROACH USING HAND GEOMETRY AND PALM PRINT TO E...
A NOVEL BINNING AND INDEXING APPROACH USING HAND GEOMETRY AND PALM PRINT TO E...
 
IRDO: Iris Recognition by fusion of DTCWT and OLBP
IRDO: Iris Recognition by fusion of DTCWT and OLBPIRDO: Iris Recognition by fusion of DTCWT and OLBP
IRDO: Iris Recognition by fusion of DTCWT and OLBP
 
Biometric key generation using crow algorithm
Biometric key generation using crow algorithmBiometric key generation using crow algorithm
Biometric key generation using crow algorithm
 
Palmprint and Handgeometry Recognition using FAST features and Region properties
Palmprint and Handgeometry Recognition using FAST features and Region propertiesPalmprint and Handgeometry Recognition using FAST features and Region properties
Palmprint and Handgeometry Recognition using FAST features and Region properties
 
Face Liveness Detection for Biometric Antispoofing Applications using Color T...
Face Liveness Detection for Biometric Antispoofing Applications using Color T...Face Liveness Detection for Biometric Antispoofing Applications using Color T...
Face Liveness Detection for Biometric Antispoofing Applications using Color T...
 
International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)
 

More from iosrjournals (20)

I0114549
I0114549I0114549
I0114549
 
H0124246
H0124246H0124246
H0124246
 
H0114044
H0114044H0114044
H0114044
 
G0145458
G0145458G0145458
G0145458
 
G0135059
G0135059G0135059
G0135059
 
G0123541
G0123541G0123541
G0123541
 
G0113839
G0113839G0113839
G0113839
 
F0144153
F0144153F0144153
F0144153
 
F0134249
F0134249F0134249
F0134249
 
F0122934
F0122934F0122934
F0122934
 
E0143640
E0143640E0143640
E0143640
 
E0133641
E0133641E0133641
E0133641
 
E0112128
E0112128E0112128
E0112128
 
D0142635
D0142635D0142635
D0142635
 
D0133235
D0133235D0133235
D0133235
 
D0121720
D0121720D0121720
D0121720
 
D0111420
D0111420D0111420
D0111420
 
C0141625
C0141625C0141625
C0141625
 
C0132131
C0132131C0132131
C0132131
 
B0140815
B0140815B0140815
B0140815
 

Recently uploaded

COMPUTER 10 Lesson 8 - Building a Website
COMPUTER 10 Lesson 8 - Building a WebsiteCOMPUTER 10 Lesson 8 - Building a Website
COMPUTER 10 Lesson 8 - Building a Websitedgelyza
 
COMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online CollaborationCOMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online Collaborationbruanjhuli
 
Designing A Time bound resource download URL
Designing A Time bound resource download URLDesigning A Time bound resource download URL
Designing A Time bound resource download URLRuncy Oommen
 
Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™Adtran
 
VoIP Service and Marketing using Odoo and Asterisk PBX
VoIP Service and Marketing using Odoo and Asterisk PBXVoIP Service and Marketing using Odoo and Asterisk PBX
VoIP Service and Marketing using Odoo and Asterisk PBXTarek Kalaji
 
Building AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptxBuilding AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptxUdaiappa Ramachandran
 
Comparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and IstioComparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and IstioChristian Posta
 
Salesforce Miami User Group Event - 1st Quarter 2024
Salesforce Miami User Group Event - 1st Quarter 2024Salesforce Miami User Group Event - 1st Quarter 2024
Salesforce Miami User Group Event - 1st Quarter 2024SkyPlanner
 
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdf
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdfUiPath Solutions Management Preview - Northern CA Chapter - March 22.pdf
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdfDianaGray10
 
How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?IES VE
 
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostKubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostMatt Ray
 
Igniting Next Level Productivity with AI-Infused Data Integration Workflows
Igniting Next Level Productivity with AI-Infused Data Integration WorkflowsIgniting Next Level Productivity with AI-Infused Data Integration Workflows
Igniting Next Level Productivity with AI-Infused Data Integration WorkflowsSafe Software
 
Videogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdfVideogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdfinfogdgmi
 
ADOPTING WEB 3 FOR YOUR BUSINESS: A STEP-BY-STEP GUIDE
ADOPTING WEB 3 FOR YOUR BUSINESS: A STEP-BY-STEP GUIDEADOPTING WEB 3 FOR YOUR BUSINESS: A STEP-BY-STEP GUIDE
ADOPTING WEB 3 FOR YOUR BUSINESS: A STEP-BY-STEP GUIDELiveplex
 
Basic Building Blocks of Internet of Things.
Basic Building Blocks of Internet of Things.Basic Building Blocks of Internet of Things.
Basic Building Blocks of Internet of Things.YounusS2
 
Linked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond OntologiesLinked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond OntologiesDavid Newbury
 
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve DecarbonizationUsing IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve DecarbonizationIES VE
 
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPAAnypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPAshyamraj55
 

Recently uploaded (20)

COMPUTER 10 Lesson 8 - Building a Website
COMPUTER 10 Lesson 8 - Building a WebsiteCOMPUTER 10 Lesson 8 - Building a Website
COMPUTER 10 Lesson 8 - Building a Website
 
COMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online CollaborationCOMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online Collaboration
 
Designing A Time bound resource download URL
Designing A Time bound resource download URLDesigning A Time bound resource download URL
Designing A Time bound resource download URL
 
Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™
 
VoIP Service and Marketing using Odoo and Asterisk PBX
VoIP Service and Marketing using Odoo and Asterisk PBXVoIP Service and Marketing using Odoo and Asterisk PBX
VoIP Service and Marketing using Odoo and Asterisk PBX
 
201610817 - edge part1
201610817 - edge part1201610817 - edge part1
201610817 - edge part1
 
Building AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptxBuilding AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptx
 
20150722 - AGV
20150722 - AGV20150722 - AGV
20150722 - AGV
 
Comparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and IstioComparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and Istio
 
Salesforce Miami User Group Event - 1st Quarter 2024
Salesforce Miami User Group Event - 1st Quarter 2024Salesforce Miami User Group Event - 1st Quarter 2024
Salesforce Miami User Group Event - 1st Quarter 2024
 
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdf
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdfUiPath Solutions Management Preview - Northern CA Chapter - March 22.pdf
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdf
 
How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?
 
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostKubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
 
Igniting Next Level Productivity with AI-Infused Data Integration Workflows
Igniting Next Level Productivity with AI-Infused Data Integration WorkflowsIgniting Next Level Productivity with AI-Infused Data Integration Workflows
Igniting Next Level Productivity with AI-Infused Data Integration Workflows
 
Videogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdfVideogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdf
 
ADOPTING WEB 3 FOR YOUR BUSINESS: A STEP-BY-STEP GUIDE
ADOPTING WEB 3 FOR YOUR BUSINESS: A STEP-BY-STEP GUIDEADOPTING WEB 3 FOR YOUR BUSINESS: A STEP-BY-STEP GUIDE
ADOPTING WEB 3 FOR YOUR BUSINESS: A STEP-BY-STEP GUIDE
 
Basic Building Blocks of Internet of Things.
Basic Building Blocks of Internet of Things.Basic Building Blocks of Internet of Things.
Basic Building Blocks of Internet of Things.
 
Linked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond OntologiesLinked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond Ontologies
 
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve DecarbonizationUsing IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
 
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPAAnypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPA
 

G0333946

  • 1. IOSR Journal of Computer Engineering (IOSRJCE) ISSN: 2278-0661 Volume 3, Issue 3 (July-Aug. 2012), PP 39-46 www.iosrjournals.org Feature Level Fusion Based Bimodal Biometric Using Transformation Domine Techniques 1 Ramachandra A C, 2Abhilash S K, 2K B Raja, 3Venugopal K R, 4L M Patnaik 1 Department of Electronics and Communication, Alpha College of Engineering, Bangalore, 560077. 2 Department of Electronics and Communication, University Visveswaraya College of Engineering, Bangalore, 3 Principal, University Visveswaraya College of Engineering, Bangalore, Karnataka, India, 4 Honorary Professor, Indian Institute of Science, Bangalore, Karnataka, India, Abstract: Bimodal biometric used to authenticate a person is more accurate compared to single biometric trait. In this paper we propose Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques (FLFBBT). The algorithm uses two physiological traits viz., Fingerprint and Face to identify a person. The Region of Interest (ROI) of fingerprint is obtained using preprocessing. The features of fingerprint are extracted using Dual Tree Complex Wavelet Transforms (DTCWT) by computing absolute values of high and low frequency components. The final features of fingerprint are computed by applying log on concatenated absolute value of high and low frequency components. The face image is preprocessed by cropping only face part and Discrete Wavelet Transforms (DWT) is applied. The approximation band coefficients are considered as features of face. The fingerprint and face image features are concatenated to derive final feature vector of bimodal biometric. The Euclidian Distance (ED) is used in matching section to compare test biometric in the database, it is observed that the values of EER and TSR are better in the case of proposed algorithm compared to individual transformation domain techniques. Keywords: Biometrics, Face, Fingerprint, DWT and DTCWT I. INTRODUCTION The authentication of a person using biometric is accurate and robust compared to traditional methods of identification. The security can be easily broken if the traditional methods of personal authentication using password and ID cards since they can be stolen or lost. In presence of sophisticated biometric identification system any unusual such as robbery, entry of an unauthorized person etc ., can be avoided. The biometric identification system is based on characteristics features of human body such as eyes, face, fingerprint, DNA etc., which are known as Physiological Biometrics which are constant throughout the life time and the behavior of person which depends on age and mood of a person such as signature, keystroke etc., known as Behavioral Biometrics, which are not constant throughout the life. The biometric system is operated in two different modes based on the application such as (i) Verification Mode: it is one to one comparison of test biometric with the template stored in database, (ii) Identification Mode: it is one to many comparisons of test with the database, if the comparison is nearer to the predefined threshold value then the person is identified. The general biometric system has data acquisition section, preprocessing section, feature extraction section, matching section. The function of data acquisition section is to collect the number of sample of biometrics in different conditions and store it as a database. In preprocessing section the images are normalized by colour converting, cropping and resizing. The features are extracted from normalized image in feature extraction section by using spatial domine, transformation domine or combination of both. The final results are obtained from extracted features in matching section by using the distance formulas like Euclidian Distance (ED), Hamming Distance, Chi-square, Linear Discriminant Analysis, Support Vector Machine, Neural Networks etc., The biometric systems are very much essential in applications like Home Security, Airport Checking, Voting machine, Entry to high security zone like parliament House, ATM, Laptop, public places like shopping mall etc., The multimodal biometrics systems address the problem of single biometric system by considering multiple evidences produced by multiple biometric traits. Multimodal biometric systems also provide anti- poofing measures by making it difficult for an intruder to spoof multiple biometric traits simultaneously. An integration fusion scheme is required to fuse the information presented by the individual modalities. Contribution: In this paper FLFBBT algorithm is proposed to identify a person. The fingerprint and face of each person is considered and features are extracted using DTCWT and DWT respectively. The ED is used in matching section to authenticate a person. Organization: This paper is organized into following sections. Section 2 is an overview of Literature Survey. The proposed model is described in Section 3. Section 4 discusses the Algorithm; Performance analysis of the system is presented in Section 5 and Conclusions are contained in Section 6. www.iosrjournals.org 39 | Page
  • 2. Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques II. LETRAERETURE SURVEY Wang Wenchao and Sun Limm [1] discussed a Fingerprint identification algorithm that combines point match and image match. This algorithm depends on singular point abstraction and wavelet transform coefficient. First singular points are extracted for rough match and image calibration. Ma Yinping and Huang Yongxing [2] introduced a kind of wavelet transform adaptive threshold of the Fingerprint image denoising method. The approach is one of the Fingerprint image wavelet decomposition, and then based on Bayes framework, selecting different optimal threshold, combined with soft threshold value method to Fingerprint image denoising, improve the Fingerprint image Peak Signal-to-Noise Ratio. Atif Iqbal and Anoop Namboodiri [3] explore the effectiveness of weak features in a cascade for filtering fingerprint databases considering a set of potential indexing features computed from minutiae triplets and minutiae quadruplets. Each stage of filtering consists of projecting the probe onto a specific line and the removal of database samples outside a window around the probe. Hao et al., [4] proposed an algorithm based on multiple minutiae partitions to match general distorted fingerprints. Minutiae in an input and a template fingerprint are globally partitioned into pairs of multiple neighbors, each of which consist of similar local minutia topology and features. The minutia similarity is additionally evaluated by investigating ridge shape and the geometric similarities of the corresponding neighbor minutiae. Jinhai Zhang [5] discussed methods of fingerprint image segmentation. Some of the segmentation methods are classified based on block variance method, Pattern methods, and synthesis method. The grading segmentation method automatically identifies some close value in a hierarchical segmentation. Arun et al., [6] proposed the feature extraction techniques for three modalities viz. Fingerprint, iris and Face. The extracted information from each modality is stored as a template. The information are fused at the match score level using a density based score level fusion, Gaussian Mixture Model (GMM) followed by the Likelihood ratio test. Ngoc-Son Vu and Alice Caplier [7] invented a feature descriptor set called Patterns of Oriented Edge Magnitudes (POEM) for feature extraction by applying the self-similarity operator on accumulated edge magnitudes across different directions. Based on these Features, two feature descriptors are constructed. Carlos et al., [8] proposed an Active Shape Models (ASM) landmark selection scheme to improve the ASM performance in face recognition applications. The proposed scheme selects robust landmark points where relevant facial features are found and assigns higher weights to their corresponding features in the face classification stage. Xiaoni Liang and Weiqing Tong [9] presented an accurate and fast approach to estimate the pose of face in near-infrared images. The approach automatically extracts the feature points from a normalized face image. The pupils and nose tip are used to estimate the pose rotation of yaw and the pupils, nose tip and the center line of mouth is used to estimate the pose rotation of pitch. Ajay Kumar and Yingbo Zhou [10] proposed a system which simultaneously acquires the finger-vein and low-resolution fingerprint images and combines these two evidences using a score-level combination strategy. The two score-level combinations are holistic and nonlinear fusion. Kolhandai et al., [11] present an intelligent hybrid features based face recognition method which combines the local and global approaches to produce a complete a robust and high success rate face recognition system. The global features are computed using principal component analysis while the local features are ascertained configuring the central moment, Eigen vectors and the standard deviation of the eyes, nose and mouth segments of the human face as the decision support entities of the Generalized Feed Forward Artificial Neural Network. Ramya et al., [12] proposed a system with key is generated from the fused feature which is of greater than 128 bit which is enough for AES encryption, Hash Encoding and AES decryption is done. Message is subjected to hash and hash of message is obtained and it is given to AES encryption as a message and the key is obtained from the fused feature. Lili Liu and Tianjie Cao [13] defined an efficient verification system based on biometrics, which access real system resources without physical key. Based on the pattern matching, the feature are extracted from the fingerprint images, the feature depends on the position and angle of the input fingerprints. The ridges and valleys structure information integration is employed to perform the fingerprint minutiae alignment. Biggio et at., [14] have investigated the robustness of different score fusion rules for multimodal biometric verification systems, against spoofing attacks for Fingerprint and Face biometrics. The simulated worst-case scenario considered is not representative of the score distribution of real spoofing attacks. Abhishek Nagar et at., [15] discussed a feature-level fusion framework for the design of multi biometric cryptosystems that simultaneously protects the multiple templates of a user using a single secure sketch. The feasibility of such a framework has been demonstrated using both fuzzy vault and fuzzy commitment. III. MODEL In this section the block diagram of the proposed model as shown in Figure 1 is discussed. The two biometric traits viz., Fingerprint and Face of a person are considered together to identify a person. The transform domain techniques such as DTCWT and DWT are used to extract Features from Fingerprint and Face respectively. www.iosrjournals.org 40 | Page
  • 3. Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques Database Section Test Section Fingerprint Face Fingerprint Face Preprocess Preprocess Preprocess Preprocess DTCWT DWT DTCWT DWT Features Features Features Features Fusion Fusion Match/Mismatch Accept/Reje ct Figure 1: Block diagram of proposed algorithm 3.1. DATABASE: 3.1.1. Fingerprint database: The DB3 Fingerprint database from FVC2004 competitions is considered to test the algorithm. The database consists of 110 persons with 8 Fingerprint images per person having image size of 300x480 with a resolution of 512 dpi. The Fingerprint images are collected in three different sessions with at least two weeks’ time separating each session using thermal sweeping sensor [16]. The DB3 has eight images of single person shown in Figure 2. The Fingerprint database is created in our proposed model by considering minimum of 10 persons and maximum of 40 persons with 6 images per persons. The number of images with database varies between 60 and 240. The Seventh and Eighth Fingerprint images of each person used in the test section. The number of persons considered with out off database varies from 5 to 30 with one Fingerprint images per person. Figure 2: Fingerprint Images of DB3 FVC2004 3.1.2. Face database: The Olivetti Research Laboratory (ORL) gray scale [17] is considered to test the proposed model. The data base has 40 persons and 10 images per person i.e., 400 Face images in the database. The each Face image has frontal view with slight tilt of head with a size of 92 x 112. To test the proposed modal, the Face data base is created by considering 10 persons to 40 persons with six images per person. The seventh and eight images of each person are considered as test images. The out off database has five persons to 30 persons with one Face image per person is used to compute FAR. The Face image samples of one person are shown in figure 3. Figure 3: Face Images of ORL www.iosrjournals.org 41 | Page
  • 4. Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques 1.2. PREPROCESSING: The Fingerprint image of size 300 x 480 is manually cropped to a dimension of 201 x 401 to obtain Region of Interest (ROI) Fingerprint area by eliminating four side edges. The size of ROI is resized to 256 x 256 to apply DTCWT for Features. The original Fingerprint image and its cropped and resized image are shown in figure 4. The original Face image of size 92 x 112 is manually cropped to obtain ROI of size 71 x 93 by eliminating edge of Face image. The ROI is resized to an arbitrary value say size of 192 x 192 to apply DWT to derive Features. The original Face image and its cropped and resized image are shown in figure 5. (a) (b) (a) (b) Figure 4: (a) Original Fingerprint Image (300 x 480) Figure 5: (a) Original Face Image (92 x 112) (b) Preprocessed Fingerprint Image (256 x 256) (b) Preprocessed Face Image (192 x 192) 1.3. FEATURE EXTRACTION: The Transform Domain Features from the Fingerprint and Face image are obtained by using the Dual Tree Complex Wavelet Transforms (DTCWT) and Discrete Wavelet Transforms (DWT) transforms respectively. The logarithm is applied on DTCWT magnitudes of Fingerprint to obtain Final Features of Fingerprint image. The Final Fingerprint Features and Features of Face are fused by concatenating to derive the Final set of Features to test the algorithm. 3.3.1. DTCWT: It produces complex coefficients by using dual tree wavelet filters and gives real and imaginary parts, which consists of both low frequency and high frequency bands. The high frequency band is ignored and low frequency band is considered and decomposed into two parts for next levels and process is continued until the required six levels as shown in Figure 6. At level six all the low frequency coefficients and high frequency coefficients from both real and imaginary parts are considered. The magnitudes of both real and imaginary parts are calculated using Equations 1 and 2 respectively and have a dimension of 1 x 48. Magnitudes are concatenated as given in Equation 3 to obtain a single Vector with a dimension of 1 x 96. For the obtained single Vector logarithmic scale is applied as given in Equation 4 to normalize the magnitude, which is considered as Feature Vector FFP for Fingerprint image. M L  [( RL 6) 2  ( I L 6) 2 ] …………………. (1) M H  [( RH 6) 2  ( I H 6) 2 ] …….……………. (2) M  [M H : M L ] …….…………….. (3) FFP  Log(M ) ……………………. (4) 3.3.2 DWT: The preprocessed Face image of size 192 x 192 is considered and one level haar wavelet is applied using base Equation 5 to derive four sub bands such as Approximation band, Horizontal band, Vertical band and Diagonal band as shown in Figure 7. The Approximation band has low frequency components and has significant information of Face image. The Horizontal, Vertical and Diagonal bands are high frequency bands and have in significant information of Face image like edge information’s. The approximation band coefficients are considered as Features of Face image FF which is of dimension 1 x 9216. Figure 6: DWT sub bands of Face image. www.iosrjournals.org 42 | Page
  • 5. Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques  1 0  t 1/ 2   (t )    1 1/ 2  t 1 ………………………… (5)  0 otherwise  1.4. FUSION OF FEATURES: The Log DTCWT feature of Fingerprint FFP and Approximation band coefficient FF of DWT of Face image are concatenated to obtain Final feature Vector set for bimodal biometrics. The Final feature Vector FFV is given in Equation six Equation 5. The Final feature Vector size is of 1x 9312. Level 6 Level 5 RL6 Real Part Level 4 RL5 Level 3 RL4 Level 2 RL3 RH Level 1 RH 6 RL2 RL1 RH 5 RH 4 RH 3 IL6 RH 2 IL5 1 IL4 IL3 IH6 X IH5 IL2 IH4 IL1 IH3 IH2 IH1 Imaginary Part Figure 6: Different levels of DTCWT. FFV  [ FFP : FF ] ………………... (6) 1.5. MATCHING: The Final fused feature Vector derived from both Fingerprint and Face images are considered to test match/non match of a person. The Final fused Features of test image are compared with Final fused Features of database images using Euclidian distance as given in Equation 7.  1 M d ( p, q)  ( pi  qi ) 2 ……………. (7) M i 1 M = Dimension of Vector. Pi = Data Feature Vector Qi = Test Feature Vector The TSR, FRR and FAR values are computed by compared ED values with predefined thresh hold values. IV. ALGORITHM Problem Definition: The proposed bimodal biometric algorithm using Log (DTCWT) and DWT to identify a person is discussed and is given in Table 1 The Objectives:  The bimodal biometric recognition is identified using fusion of DTCWT and DWT coefficients.  To reduce FAR,FRR and EER  To increase TSR www.iosrjournals.org 43 | Page
  • 6. Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques Table 1: Algorithm of proposed System. INPUT: Fingerprint and Face images. OUTPUT: Matched /Non-matched Test image 1. Resize Fingerprint image to 256x256. 2. DTCWT is applied on Fingerprint of database and test Fingerprint. 3. The Fingerprint initial features are obtained by concatenation of magnitudes of low and high frequency components. 4. The Fingerprint final features are derived by Log of initial features. 5. Resizing of Face image to 192x192. 6. DWT is applied on database images and test image. 7. The Face features are obtained by considering coefficients of approximate bands. 8. The bimodal biometric final features are derived by concatenation of Fingerprint and Face features. 9. The final features of test image are compared with final features of images in the database using ED. V. PERFORMANCE ANALYSIS: The Fingerprint databaseFVC2004 DB3 and Face database ORL are considered to test the proposed algorithm in terms of the following performance parameters. 1. False Acceptance Rate (FAR): It is the probability that an unauthorized person is incorrectly accepted as authorized person and it is computed using Equation 8 FAR = Number of unauthorized persons Identified Total Number of unauthorized persons ………………………… (8) 2. False Rejection Rate (FRR): It is the probability that the authorized person is incorrectly rejected and it is computed using the Equation 9. FRR = Number of authorized persons rejected Total Number of authorized persons ……………………….. (9) 3. Equal Error Rate (EER): It is the rate at which both FAR and FRR are equal. 4. Total Success Rate (TSR): It is the rate at which match occurs successfully. The Numbers of persons recognized correctly in the database is given in the Equation 10. %TSR = Number of persons recognized correctly * 100 Total Number of persons in the database …………………………. (10) The proposed algorithm is robust in recognition of a person as it uses two biometric traits. The variation of FRR, FAR and percentage TSR with threshold is tabulated in the Table 2 As threshold value increases the FRR values decreases from 1 to zero whereas FAR and TSR values increases .The maximum percentage TSR value of 90% is achieved from 0.7 threshold value. The variation of EER and TSR values for test images 7 and 8 of each person with Persons Inside Database (PIDB): Persons Outside Database (PODB) are given in Table 3.The values of EER increases with PODB and PIDB.The values of EER and TSR depends on the test images, i.e., for test image number 7, EER is 0.16 and TSR is 70% for PIDB and PODB combination of 10:30 and for test image number 8, EER is 0.13 and TSR is 90%. For PIDB and PODB combination10:30. Table 2: FRR, FAR and % TSR for different Threshold. Thresh Hold FRR FAR %TSR 0 1 0 0 0.05 1 0 0 0.25 1 0 0 0.3 0.9 0 10 0.4 0.8 0 20 0.45 0.6 0 40 0.5 0.5 0 50 www.iosrjournals.org 44 | Page
  • 7. Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques 0.55 0.4 0.033 60 0.6 0.2 0.033 70 0.65 0.2 0.1 70 0.7 0 0.2 90 0.75 0 0.4 90 0.8 0 0.633 90 0.85 0 0.733 90 0.9 0 0.833 90 0.95 0 0.9 90 1 0 0.96 90 Table 3: Comparison of EER and Percentage of TSR for different PIDB: PODB. The variations of FAR and FRR with threshold for PIDB and PODB combination 10:30 is given in Figure 7. As threshold increases FRR decreases and FAR increases the EER value of 0.13 is intersection of FRR and FAR plot. Figure 7: Plot of FAR and FRR V/S Thresh Hold with Log The values of EER and TSR for DTCWT, DWT and combinations of LOG (DTCWT) and DWT i e., proposed algorithm is given in Table 4. It is observed that the proposed algorithm has low EER value compared to individual DTCWT and DWT algorithms. The value of percentage TSR is high in the case of proposed algorithm compared to individual transformation algorithms. Hence the proposed bimodal biometric algorithm has better efficiency and also robust as two biometric traits are used to identify a person compared to single biometric trait. Table 4: Comparison of EER and Percentage of TSR for different Transforms and Proposed Algorithm. Method Parameters DTCWT DWT Proposed FLFBBT EER 0.18 0.485 0.13 %TSR 80 40 90 VI. CONCLUSION: In this paper FLFBBT algorithm is proposed to identify a person accurately. The DTCWT is applied on ROI of fingerprint and computed magnitudes of low and high frequency components. The features of fingerprint are derived by applying log on concatenated magnitudes of low and high frequency components. The DWT is applied on face image and features are extracted by coefficients of approximate band. The final features of bimodal biometrics are computed by concatenating fingerprint and face features. The ED is used for comparison to authenticate a person. It is observed that the values of EER and TSR are better in the proposed algorithm compared to individual transform domain techniques. In future the algorithm can be verified by using both spatial domain techniques and transform domain techniques at different levels of fusion. www.iosrjournals.org 45 | Page
  • 8. Feature Level Fusion based Bimodal Biometric using Transformation Domine Techniques REFERENCES [1] Wang Wenchao and Sun Limin, “A Fingerprint Identification Algorithm Based on Wavelet Transformation Characteristic Coefficient,” International Conference on Systems and Informatics, pp1-3, 2012. [2] Ma Yinping and Huang Yongxing, “Adaptive Threshold Based on Wavelet Transform Fingerprint Image Denoising,” International Conference on Computer Science and Electronics Engineering, pp. 494- 497, 2012. [3] Atif Iqbal and Anoop Namboodiri, “Cascaded Filtering for Fingerprint Identification using Random Projections,” Computer Society Conference on Computer Vision and Pattern Recognition Workshops, pp. 77- 82, 2012. [4] Hao NI, Dongju LI, Tsuyoshi Isshiki and Hiroaki Kunieda, “Robust Multiple Minutiae Partitions for Fingerprint Authentication,” International Symposium on Biometrics and Security Technologies, pp. 35- 44, 2012. [5] Jinhai Zhang, “The research of fingerprint image Segmentation method,” Second International Conference on Consumer Electronics, Communications and Networks, pp. 701- 704, 2012. [6] S ArunVivek, J Aravinth and S Valarmathy, “Feature Extraction for Multimodal Biometric and Study of Fusion Using Gaussian Mixture Model,”Proceedings of the International Conference on Pattern Recognition, Informatics and Medical Engineering, pp. 391-392, 2012. [7] Ngoc-Son Vu and Alice Caplier, “Enhanced Patterns of Oriented Edge Magnitudes for Face Recognition and Image Matching” IEEE Transactions on Image Processing, vol. 21, no. 3, 2012. [8] Carlos A. R. Behaine and Jacob Scharcanski, “Enhancing the Performance of Active Shape Models in Face Recognition Applications,” IEEE Transactions on Instrumentation and Measurement, vol. 61, no. 8, 2012. [9] Xiaoni Liang and Weiqing Tong, “Face Pose Estimation Using Near-Infrared Images,” International Conference on Communication Systems and Network Technologies, pp. 216- 220, 2012. [10] Ajay Kumar and Yingbo Zhou, “Human Identification Using Finger Images,” Transactions on Image Processing, vol. 21, no. 4, April 2012. [11] Kolhandai Yesu, Himadri Jyoti Chakravorty, Prantik Bhuyan, Rifat Hussain and Kaustubh Bhattacharyya, “Hybrid Features Based Face Recognition Method Using Artificial Neural Network,” Second National Conference on Computational Intelligence and Signal Processing, pp. 40 – 46, 2012. [12] M.Ramya, A. Muthukumar and S.Kannan, “Multi biometric based Authentication using Feature Level Fusion,” International Conference on Advances In Engineering, Science And Management, pp. 191- 195, 2012. [13] Lili Liu and Tianjie Cao, “The Research and Design of an Efficient Verification System Based on Biometrics,” International Conference on Computer Science and Electronics Engineering, pp. 707-710, 2012. [14] B Biggio, Z Akhtar, G Fumera, G L Marcialis and F Roli, “Security Evaluation of Biometric Authentication Systems Under Real Spoofing attacks,” IET Biometrics, Vol. 1, pp. 11–24, 2012. [15] Abhishek Nagar, Karthik Nandakumar and Anil K Jain, “Multibiometric Cryptosystems Based on Feature Level Fusion,” IEEE Transactions on Information Forensics and Security, vol. 7, no.1, pp. 225-268, 2012. [16] http://bias.csr.unibo.it/fvc2004/default.asp, FVC2004-Third International Fingerprint Verification Competition. [17] Ferdinando Samaria and Andy Harter, “Parameterization of a Stochastic Model for Human Face Identification,” Proceedings of Second IEEE Workshop on Applications of Computer Vision, December 1994. www.iosrjournals.org 46 | Page