ArticlePDF Available

Review Paper on Applications of Principal Component Analysis in Multimodal Biometrics System

Abstract and Figures

Unimodal biometric systems are susceptible to a variety of problems such as noisy data, intra-class variations, limited degrees of freedom, non-universality, spoof attacks and unacceptable error rates. Some of these limitations can be addressed by deploy multimodal biometric systems that integrates the evidence presented by multiple sources of information The proposed system provides effective fusion scheme that combines information presented by the multiple domain experts based on the Rank level fusion integration method, thereby increasing the efficiency of the system which is not possible by the unimodal biometric system. The proposed multimodal biometric system has a number of unique qualities, including principal component analysis and fisher's linear discriminate methods for individual matchers authentication. The novel rank level fusion method is used in order to consolidate the results obtained from different biometric matchers.
Content may be subject to copyright.
Procedia Computer Science 92 ( 2016 ) 481 486
Available online at
1877-0509 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
Peer-review under responsibility of the Organizing Committee of ICCC 2016
doi: 10.1016/j.procs.2016.07.371
2nd International Conference on Intelligent Computing, Communication & Convergence
Srikanta Patnaik, Editor in Chief
Conference Organized by Interscience Institute of Management and Technology
Bhubaneswar, Odisha, India
Review paper on applications of principal component analysis in
multimodal biometrics system
Chhaya Sunil Khandelwal a, Ranjan Maheshewarib, U.B.Shinde
a Department of Electronics and Telecommunication , Jawaharlal Nehru Engineering College, Aurangabad,431005 , India.
b Department of Electronics ,Professor, RTU, Kota(Rajasthan), 324010, India.
*Department of Electronics and Telecommunication, Principal CSMSSCOE, Aurangabad, 431005, India.
Unimodal biometric systems are susceptible to a variety of problems such as noisy data, intra-class variations,
limited degrees of freedom, non-universality, spoof attacks and unacceptable error rates. Some of these limitations
can be addressed by deploy multimodal biometric systems that integrates the evidence presented by multiple
sources of information The proposed system provides effective fusion scheme that combines information presented
by the multiple domain experts based on the Rank level fusion integration method, thereby increasing the efficiency
of the system which is not possible by the unimodal biometric system. The proposed multimodal biometric system
has a number of unique qualities, including principal component analysis and fisher’s linear discriminate methods
for individual matchers authentication. The novel rank level fusion method is used in order to consolidate the results
obtained from different biometric matchers.
© 2016 The Authors. Published by Elsevier B.V.
Selection and peer-review under responsibility of scientific committee of Interscience Institute of Management and Technology.
Keywords: Multimodal biometrics; PCA; Fingerprint Recognition; Iris Recognition; Minutiae Extraction.
* Corresponding author. Tel.: 7588402678; 9460813252;90911191795 ,fax: +0-000-000-0000 .
E-mail, ,
© 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
Peer-review under responsibility of the Organizing Committee of ICCC 2016
482 Chhaya Sunil Khandelwal et al. / Procedia Computer Science 92 ( 2016 ) 481 – 486
1. Introduction
Biometrics has drawn wide acceptability during the last 35 years. It is used for building and store access control,
image identification, observation and computer interfacing. The key issue of these applications is the identification
of individuals by their physiological or behavioral characteristics (e.g., face, fingerprint, iris, signature, or
gait)[1][2]. Each biometric characteristic has its own strengths and weaknesses: but, none is free from any one or
more issues such as noisy data, non universality, spoof attacks, and unacceptable error rates. In the past few years,
researchers have more and more focused on the possibility of including multiple sources of information.
A simple biometric system consists of four basic components: A simple biometric system consists of four basic
1) Sensor module which acquires the biometric data;
2) Feature extraction module where the acquired data is processed to extract feature vectors;
3) Matching module where feature vectors are compared against those in the template;
4) Decision-making module in which the user's identity is established or a claimed identity is accepted or rejected.
Any human physiological or behavioural trait can serve as a biometric characteristic as long as it satisfies the
following requirements:[1][2][8]
1) Universality. Everyone should have it, barring a few exceptions, like physical deformities;
2) Distinctiveness. No two individuals should have the same characteristics;
3) Permanence. It should be invariant over a given period of time;
4) Collectability. The feature should be sensed the given system.
2. Literature survey
Multimodal techniques are not new to the medical world. In routine medical checkups also, it is often preferred
have a primary and a confirmatory examination. The inclusion of evidences from more than one sources would
enhance the overall Accuracy of the system.
Table 1. Literature Survey.
Biometric Modalities
Level of fusion
Accuracy reported
Vincenzo Conti et al. [3]
Fingerprint and iris
Feature level fusion
Abhishek Nagar et al. [4]
Iris, fingerprint and face
Feature level fusion
Robert Snelick et al. [5]
fingerprint, face
Simple-Sum fusion
A. Muthukumar et al. [6]
Iris and fingerprint
Score fusion
Sumit Shekhar, et al. [7]
Iris, fingerprint and face
Sparse matrices
3. Biometric system errors
The Biometrics signal acquisition is not free from errors. When errors are significant, two samples of the same
biometric characteristic from the same subject (e.g., two impressions of a user’s right index finger) may not exactly
be the same due to imperfect imaging conditions (e.g., sensor noise and dry fingers), changes in the user’s
physiological or behavioural characteristics (e.g., cuts and bruises on the finger), ambient conditions (e.g.,
temperature and humidity), and user’s interaction with the sensor (e.g.finger placement). Therefore, the response of
Chhaya Sunil Khandelwal et al. / Procedia Computer Science 92 ( 2016 ) 481 – 486
Fig.1. Biometric system error rates
The main system errors are usually measured in terms of:
FNMR (false no match rate) mistaking two biometrics measurements from the same person to be from two
different persons;
FMR (false match rate) mistaking biometric measurement from two different persons to be from the same
Multimodal biometric systems can be designed to operate in five integration scenarios: 1)multiple sensors, 2)
multiple biometrics, 3) multiple units of the same biometric, 4) multiple snapshots of the same biometric, 5) multiple
representations and matching algorithms for the same biometric[1][4][8]
4. Fusion in multimodal biometrics
A biometric system has four important modules. The sensor module acquires the biometric data from a user; the
feature extraction module processes the acquired biometric data and extracts a feature set to represent it; the
Decision Threshold
484 Chhaya Sunil Khandelwal et al. / Procedia Computer Science 92 ( 2016 ) 481 – 486
matching module compares the extracted feature set with the stored templates using a classifier or matching
algorithm in order to generate matching scores; in the decision module the matching scores are used either to
identify an enrolled user or verify a user’s identity. Sanderson and Paliwal [10] have classified information fusion in
biometric systems into two broad categories: pre-classification fusion and post-classification fusion. Pre-
classification fusion refers to combining information prior to the application of any classifier or matching algorithm.
In post-classification fusion, the information is combined after the decisions of the classifiers have been obtained.
The figure (2) shows the block diagram of unimodal and multimodal biometric system. In a unimodal biometric
system, the first stage is the enrollment stage , next is the feature extraction stage in which the features of each
person to be identified is extracted and next is the feature matching stage in which the obtained features are
compared with the data base and the output is obtained.[11] In a multimodal biometric system , after the enrollment
stage the images are averaged and normalized and then its given to the matching stage in which the features are
matched and then it is ranked according to the availability of data and the result is obtained.
(b) Fig.2.Block Diagram of (a) Unimodal and (b)Multimodal Biometric System
4.1 Principal component analysis
Principal component analysis (PCA) is one of the statistical techniques frequently used in signal Processing to the
dimension reduction or to the data decorrelation.PCA takes the advantage of Eigenvectors properties for
determination of selected object orientation. PCA belongs to linear Transforms based on the statistical techniques.
This method provides a powerful tool for data analysis and pattern recognition which is often used in signal and
image processing as a technique for data compression, data dimension reduction or their de correlation as well. PCA
Chhaya Sunil Khandelwal et al. / Procedia Computer Science 92 ( 2016 ) 481 – 486
is a statistical method which involves analysis of n-dimensional data. PCA observes correspondence between
different dimensions and determines principal dimensions, along which the variations of the data is high. The basis
dimensions or vectors computed by PCA are in the directions of the largest variance of the training vectors. These
basis vectors are computed by solution of an “Eigen” problem, the basis vectors are eigenvectors. The Eigen vectors
are defined in the image space. They can be viewed as images. Hence, they are usually referred to as Eigen images.
Eigen image recognition derives its name from the German prefix Eigen, meaning own or individual. The first Eigen
image is the average image, while the rest of the Eigen images represent variations from this averaging image. When
a particular image is projected onto the image space, its vector (made up of its weighted values with respect to each
Eigen image) into the image space describes the importance of each features in the image. In our system the Eigen
image approach is used, because it has several advantages. In the context of personal Identification, the background
transformations can be controlled and the Eigen image approach has a compact representation of an image can be
concisely represented by a feature vector with a few elements. Also, it is feasible to index an Eigen image-based
template database using different indexing techniques such as retrieval can be conducted efficiently. Moreover, the
Eigen image approach is a generalized template-matching approach which was demonstrated to be more accurate
than the attribute based approach.[2][12]
4.2 Rank level fusion approach
Rank-level fusion is a relatively new fusion approach. When the output of each biometric matcher is a subset of
possible matches sorted in decreasing order of confidence, fusion can be done at the rank level. The goal of rank-
level fusion is to consolidate the rank output by individual biometric subsystems (matchers) in order to derive a
consensus rank for each identity using three methods to combine the ranks assigned by different matchers[13].
1. The Highest Rank Method,
2. The Borda Count Method and
3. The Logistic Regression Method.
In the highest rank method, each possible match is assigned the highest (minimum) rank, as computed by different
matchers. The Borda count method uses the sum of the ranks assigned by individual matchers to calculate the final
rank.. In the logistic regression method, a weighted sum of the individual ranks is calculated. The weight to be
assigned to different matchers is determined by logistic regression.[6][14]
5. Conclusion
The multimodal biometrics is a promising area of information processing research which is directed towards
understanding of traits and methods for more accurate and reliable personal information representation for
subsequent decision making and matching. In the recent years there is a significant increase in research activity
directed at understanding all aspects of biometric information system representation and utilization for decision-
making support, for use by public and security services, medical diagnostics, and for understanding the complex
processes behind biometric matching and recognition.
The Authors are indebted to their respective institutions and also to the National Institute of Electronics and
Information Technology, Aurangabad, where the first author is a registered for her research work and the other
authors are registered as supervisors. Thanks to Ranjan Maheshwari sir who provided valuable VXJJHVWLRQVWR
486 Chhaya Sunil Khandelwal et al. / Procedia Computer Science 92 ( 2016 ) 481 – 486
1. Kresimir Delac 1, Mislav Grgic 2, A survey of biometric recognition methods , 46th International Symposium Electronics in Marine,
Elmar-2004, 16-18 June 2004; Zadar, Croatia
2. Anil Jaina, Karthik Nandakumara, Arun Rossb, Score normalization in multimodal biometric systems, Pattern Recognition 38 (2005) 2270
2285, accepted18 January 2005
3. Vincenzo Conti, Carmelo Militello, Filippo Sorbello, Member, IEEE, and Salvatore Vita bile, Member, IEEE, A Frequency-based
Approach for Features Fusion in Fingerprint and Iris Multimodal Biometric Identification Systems, IEEE Tractions on system, Man, and .
cybernetics part C: Application and Reviews, vol. 40, no. 4, July 2010
4. Abhishek K. Nagar, Student Member, IEEE, Karthik Nandakumar, Member, IEEE, and Anil K. Jain, Fellow, IEEE ,Multibiometric
Cryptosystems Based on Feature-Level Fusion, IEEE Tractions on Information Forensics and Security, vol.7, no.1, February 2012.
5. Robert Snelick1, Umut Uludag2*, Alan Mink1, Michael Indovina1 and Anil Jain2 Large Scale Evaluation of Multimodal Biometric
Authentication Using State-of-the-Art Systems IEEE Transactions on Pattern Analysis and Machine Intelligence Vol. 27, No. 3, Mar
2005, pp 450-455.
6. A. Muthukumar1, C. Kasthuri2 and S. Kannan3,Multimodal Biometric Authentication using Particle Swarm Optimization Algorithm with
Fingerprint and Iris ,ICTACT Journal on Image and video processing, February 2012, volume: 02, Issue: 03
7. Sumit Shekhar, Student Member, IEEE, Vishal M. Patel, Member, IEEE Nasser M. Nasrabadi, Fellow,IEEE, and Rama Chellappa, Fellow,
IEEE,Joint Sparse Representation for Robust Multimodal Biometrics, IEEE Tractions on pattern analysis and machine Intelligence, vol.
36, no. 1,January 2014.
8. Anil K. Jain, Fellow, IEEE, Arun Ross, Member, IEEE, and Salil Prabhakar, Member, IEEE, An Introduction to Biometric Recog nition,
IEEE Tranctions on circuits and systems for video Technology,Vol.14,No.1,January 2004.
DQG,ULV International Journal of Computer Science and Information Security,Vol. 7, No. 2, February 2010
10. C. Sanderson, K.K. Paliwal, Information fusion and person verification using speech and face information, Research Paper ID IAP-RR 02-
33, IDIAP, September 2002.
11. U. M. Bubeck and D. Sanchez, Biometric authentication: Technology and evaluation, San Diego State Univ., San Diego, CA, 2003. Tech. 9.
12. Md.Maruf Monwar , and Marina L. Gavrilova, Multimodal biometric system using Rank Level Fusion Approach, IEEE Transactions on
System, Man and Cybernetics Part B: Cybernetics, Vol.39, No.4, August 2009.
13. M. P. Down and R. J. Sands, “Biometrics: An overview of the technology, challenges and control considerations,” Inf. Syst. Control J., vol.
4, 2004; pp. 53 56
14. '0DOWRQL'0DLR$.-DLQDQG63UDEKDNDUHandbook of FingerprintRecognition1HZ<RUN6SULQJHU9HUODJ
15. S. Prabhakar, S. Pankanti, A. K. Jain,Biometric Recognition: Security and Privacy Concerns, IEEE Security & Privacy, March/April
2003; pp. 33-42.
... The literature prevalently indicates that it is possible to practically apply biometric data which have the following characteristics (Gaikwad & Pasalkar, 2004;Khandelwal et al., 2016;Kumar & Farik, 2016): − Universality. Practically all people are a source of such data, except for those with congenital defects or disabilities acquired throughout life. ...
... It should be underscored that there is no single ideal and universal biometric technique. Each solution has its advantages and disadvantages which determine whether it may be used in particular areas (Khandelwal et al., 2016). What can prove to be a partial answer to the issue of seeking the ideal solution is biometric system fusion (Jagadiswary & Saraswady, 2016;Lumini & Nanni, 2017). ...
Full-text available
Research background: The ongoing digitisation process in the banking sector, coupled with widespread remote provision of services, is leading to the advent of new solutions in the field of broadly understood security. The increasingly sophisticated forms of attacks on banks’ IT systems and their users have engendered the need to implement authentication methods that would ensure high security levels, but would also be convenient for banks' clients and suited to the requirements of mass service. Biometric technology seems to be a solution to this issue. The two factors that may boost its proliferation are: the fact that banks need to adjust to more rigid regulations, and technological advancements leading to cost reduction and increased availability of biometric solutions in mobile devices. Purpose of the article: The purpose of the article is to assess the prospects for the application of biometrics by the banking sector in Poland in individual customer service channels. Methods: The basis for theoretical considerations comprises the analysis of literature on authentication techniques and research on the processes whereby consumers accept new technologies. The empirical part of the paper was based on the results of the authors’ questionnaire survey among representatives of the Polish banking sector. Findings & value added: The banking sector in Poland is on the verge of a sweeping bio-metric revolution in the coming years, because most traditional identity verification and authorisation methods currently available in banking do not comply with strict security and user convenience requirements and RTS regulations. Biometric technologies will be of use in all customer service channels, with the experts indicating authentication in bank branches, ATMs, and mobile banking as the primary implementation areas. Solutions which are most likely to be applied are those based on biometrics of fingerprints and finger veins, as well as voice biometrics.
... In the last 35 years, extensive acceptability has been attained by the BMs [1] as they are capable of detecting individuals utilizing behavioural features like handwriting and voice, physical ones, for instance, FP and retina [2,3]. As of individuals, patterns along with templates are extracted to attain real-time BM recognition; then, the extracted ones are matched with the enrolled ones; finally, the authentication decision will be accepted or rejected [4]. ...
Full-text available
Unimodal Bio-Metric (BM) systems are vulnerable to changes in an individual’s BM features in addition to presentation attacks; thus, in identifying individuals, they possess only limited effectiveness. For attaining higher dependability of BM authentication, a multi-modal BM has been implemented in authentication systems. By utilizing a Rule-based Adaptive Neuro-Fuzzy Inference System (R-ANFIS), an effectual Feature-Score-Rank (FSR) fusion-centered Multi-Modal Biometric Authentication (MMBA) has been proposed here. Face, eye, and Fingerprint (FP) are the ‘3’ images, which are taken as of the person's SDUMLA-HMT database, regarded as input in MMBA. In the proposed methodology, (i) Face image segmentation utilizing Improved Viola-Jones Algorithm, (ii) Feature Extraction (FE) to form the segmented facial parts, and (iii) Feature Selection (FS) utilizing Chaos-based Salp Swarm Algorithm (CSSA) are the operations performed on the inputted face image after gathering the input data. After that, by means of the local miniature FE along with CSSA-FS, the FP image is processed. Next, by utilizing Canny Edge-centric Modified Circular Hough Transform and CSSA-FS, the eye image is processed via iris part segmentation. Subsequently, the chosen features of ‘3’ inputted images are fused in the sequence of the FSR level. Lastly, for identifying if the person is an authentic one or not, these fused features are inputted into the R-ANFIS. Then, experimentations are conducted to evaluate the proposed methodology’s performance. The experiential outcomes displayed that when analogized with the prevailing algorithms, the proposed model achieves superior performance.
... In general, recognition of the person's identity is crucial in a wide variety of applications to render services only to genuine users. Nowadays, because of the advantage of reliability and stability, biometric technologies [10] have rapidly developed to solve this problem. Basically, this technology uses biometric characteristics such as the behavioral characteristics (e.g., voice, signature, gait, keystroke dynamics, etc.), physiological characteristics (e.g., face, palmprint, fingerprint, iris, etc.) and biological characteristics (e.g., DNA, retina, etc.) of the person to determine precisely his identity. ...
... Although biometric systems have their limitations, they have an advantage over traditional security methods in which it is very difficult to lose, steal or fragment biometric features; In addition, they facilitate human recognition at a distance [22]. Biometric systems also present a practical aspect for the user that may not be possible using traditional security techniques [23]. Users who keep different passwords for different applications may have difficulty remembering the password associated with a specific application [24]. ...
The rapid growth in the development of biometric applications causes human identification as an important issue. In addition fingerprint as a part of biometrics leading to an efficient method of human identification. Many problems appears in fingerprint patterns such as noisy patterns, confused patterns, unclear patterns displacement of patterns, spread of ink … etc. The main objective of this paper is to design and implement an efficient and effective approach for human identification using human fingerprint. The main operation of this approach after preprocessing is to localize and recognize the minutiae in fingerprint image. This approach depends on the thinning operation that is so important to prepare the image to recognize the minutiae. Good results have been achieved via implementing this approach, where the obtained similarity ratio is approximately 90%.
... Matching module: also called as decision making module that measures the similarity of features during recognition with a reference template by applying various methods like probabilistic measures, neural networks, etc. Decision module maintains the expressive templates for enrolled users as a reference for future comparison in a central database. Identified features are compared by the system with the stored reference to establish whether claimed identity is accepted or rejected [20]. Performance parameters used for verification of biometric are Equal Error Rate (EER): Predetermination of threshold values for its false acceptance rate and its false rejection rate, FAR = FRR, also called as crossover error rate (CER). ...
... This reduction is achieved by transforming to a new set of variables, the principal components, which are uncorrelated, and which are ordered so that the first few retain most of the variation present in all of the original variables (Jolliffe 2002). PCA is used in many fields of science, e.g., in digital image compression (Ng 2017), in biometrics (Khandelwal, Maheshewari, and Shindec 2016) and in agriculture (Rotaru et al. 2012). Principal components analysis is a tool for exploring data and is widely used to organize data by their respective grouping and graphical presentation (Wolds, Esbensen, and Geladi 1987). ...
This research deals with the effects of soil amendment with sewage sludge compost (SSC) on long–term quantitative changes of available microelements [iron (Fe), zinc (Zn), manganese (Mn), copper (Cu)] in two types of soils using various extraction solutions [diethylenetriaminepentaacetic acid (DTPA) and hydrochloric acid (HCl)]. These changes were visualized using principal component analysis (PCA) and hierarchical cluster analysis. A 3-year pot experiment was conducted in a randomized, factorial design with two soils (light vs. medium) and a treatment with SSC at the rate of 6 Mg ha⁻¹. The application of SSC to the light soil increased the HCl-extractable Cu, Zn, Mn, and Fe. The effect of SSC application on the microelement quantitative changes was lower in the medium soil, as shown primarily by DTPA extraction. The results of the PCA highlighted that one can correctly conclude on the basis of the limited collection of data without loss of information, which is incorporated by individual microelements.
Full-text available
We propose in this paper a multimodal biometrics system for feature extraction of iris and fingerprint. That basic objective of a biometric detection system is to automatically discriminate between subjects in a reliable manner, Multimodal biometric detection systems aim to provide false acceptance rate (FAR) and false rejection rate (FRR) for two or more physical or behavioral traits.The proposed multimodal biometric system has used principal component analysis and fisher's linear discriminate methods for individual iris matchers authentication and minutia for fingerprint and the novel rank level fusion method is used in order to consolidate the results obtained from.
The last few years have plunged us at high speed into a new means of communication, namely the Internet, which has set a new trend for the next millennium. So, the rapid growth of online applications reflects the speed with which most countries can develop. An essential aspect of online communication is related to the trust of users and is a very necessary element to ensure the success of an online application. One of the main elements underlying this trust is the remote authentication of the user through its biometric features while of course protecting these features in different storage media. In this paper, we propose a new palmprint/palm-vein recognition framework based on a hand-craft image feature learning method is suggested. Furthermore, to increase the anti-spoof capability of the system, an effective biometric templates protection method based on chaotic systems was proposed. Experimental results have shown that high accuracy can be obtained with a very high level of template protection, which implies that the proposed cancelable biometric system can operate in highly secure applications.
Full-text available
Traditional biometric recognition systems rely on a single biometric signature for authentication. While the advantage of using multiple sources of information for establishing the identity has been widely recognized, computational models for multimodal biometrics recognition have only recently received attention. We propose a multimodal sparse representation method, which represents the test data by a sparse linear combination of training data, while constraining the observations from different modalities of the test subject to share their sparse representations. Thus, we simultaneously take into account correlations as well as coupling information among biometric modalities. A multimodal quality measure is also proposed to weigh each modality as it gets fused. Furthermore, we also kernelize the algorithm to handle nonlinearity in data. The optimization problem is solved using an efficient alternative direction method. Various experiments show that the proposed method compares favorably with competing fusion-based methods.
Full-text available
The basic aim of a biometric identification system is to discriminate automatically between subjects in a reliable and dependable way, according to a specific-target application. Multimodal biometric identification systems aim to fuse two or more physical or behavioral traits to provide optimal False Acceptance Rate (FAR) and False Rejection Rate (FRR), thus improving system accuracy and dependability. In this paper, an innovative multimodal biometric identification system based on iris and fingerprint traits is proposed. The paper is a state-of-the-art advancement of multibiometrics, offering an innovative perspective on features fusion. In greater detail, a frequency-based approach results in a homogeneous biometric vector, integrating iris and fingerprint data. Successively, a hamming-distance-based matching algorithm deals with the unified homogenous biometric vector. The proposed multimodal system achieves interesting results with several commonly used databases. For example, we have obtained an interesting working point with FAR = 0% and FRR = 5.71% using the entire fingerprint verification competition (FVC) 2002 DB2B database and a randomly extracted same-size subset of the BATH database. At the same time, considering the BATH database and the FVC2002 DB2A database, we have obtained a further interesting working point with FAR = 0% and FRR = 7.28% ÷ 9.7%.
Full-text available
Multibiometric systems are being increasingly deployed in many large-scale biometric applications (e.g., FBI-IAFIS, UIDAI system in India) because they have several advantages such as lower error rates and larger population coverage compared to unibiometric systems. However, multibiometric systems require storage of multiple biometric templates (e.g., fingerprint, iris, and face) for each user, which results in increased risk to user privacy and system security. One method to protect individual templates is to store only the secure sketch generated from the corresponding template using a biometric cryptosystem. This requires storage of multiple sketches. In this paper, we propose a feature-level fusion framework to simultaneously protect multiple templates of a user as a single secure sketch. Our main contributions include: 1) practical implementation of the proposed feature-level fusion framework using two well-known biometric cryptosystems, namely, fuzzy vault and fuzzy commitment, and 2) detailed analysis of the trade-off between matching accuracy and security in the proposed multibiometric cryptosystems based on two different databases (one real and one virtual multimodal database), each containing the three most popular biometric modalities, namely, fingerprint, iris, and face. Experimental results show that both the multibiometric cryptosystems proposed here have higher security and matching performance compared to their unibiometric counterparts.
Full-text available
In many real-world applications, unimodal biometric systems often face significant limitations due to sensitivity to noise, intraclass variability, data quality, nonuniversality, and other factors. Attempting to improve the performance of individual matchers in such situations may not prove to be highly effective. Multibiometric systems seek to alleviate some of these problems by providing multiple pieces of evidence of the same identity. These systems help achieve an increase in performance that may not be possible using a single-biometric indicator. This paper presents an effective fusion scheme that combines information presented by multiple domain experts based on the rank-level fusion integration method. The developed multimodal biometric system possesses a number of unique qualities, starting from utilizing principal component analysis and Fisher's linear discriminant methods for individual matchers (face, ear, and signature) identity authentication and utilizing the novel rank-level fusion method in order to consolidate the results obtained from different biometric matchers. The ranks of individual matchers are combined using the highest rank, Borda count, and logistic regression approaches. The results indicate that fusion of individual modalities can improve the overall performance of the biometric system, even in the presence of low quality data. Insights on multibiometric design using rank-level fusion and its performance on a variety of biometric databases are discussed in the concluding section.
Full-text available
We examine the performance of multimodal biometric authentication systems using state-of-the-art Commercial Off-the-Shelf (COTS) fingerprint and face biometric systems on a population approaching 1,000 individuals. The majority of prior studies of multimodal biometrics have been limited to relatively low accuracy non-COTS systems and populations of a few hundred users. Our work is the first to demonstrate that multimodal fingerprint and face biometric systems can achieve significant accuracy gains over either biometric alone, even when using highly accurate COTS systems on a relatively large-scale population. In addition to examining well-known multimodal methods, we introduce new methods of normalization and fusion that further improve the accuracy.
Full-text available
Various aspects and advantages of biometric system are presented. A biometric system is essentially a pattern-recognition system that recognizes a person based on a feature vector derived from a specific physiological or behavioral characteristic that the person possesses. Depending on the application context, a biometric system typically operates in one of two modes: verification or identification. In verification mode, the system validates a person's identity by comparing the captured biometric characteristic with the individual's biometric template, which is prestored in the system database.
In general, the identification and verification are done by passwords, pin number, etc., which is easily cracked by others. In order to overcome this issue biometrics is a unique tool for authenticate an individual person. Nevertheless, unimodal biometric is suffered due to noise, intra class variations, spoof attacks, non-universality and some other attacks. In order to avoid these attacks, the multimodal biometrics i.e. combining of more modalities is adapted. In a biometric authentication system, the acceptance or rejection of an entity is dependent on the similarity score falling above or below the threshold. Hence this paper has focused on the security of the biometric system, because compromised biometric templates cannot be revoked or reissued and also this paper has proposed a multimodal system based on an evolutionary algorithm, Particle Swarm Optimization that adapts for varying security environments. With these two concerns, this paper had developed a design incorporating adaptability, authenticity and security.
Multimodal biometric systems consolidate the evidence presented by multiple biometric sources and typically provide better recognition performance compared to systems based on a single biometric modality. Although information fusion in a multimodal system can be performed at various levels, integration at the matching score level is the most common approach due to the ease in accessing and combining the scores generated by different matchers. Since the matching scores output by the various modalities are heterogeneous, score normalization is needed to transform these scores into a common domain, prior to combining them. In this paper, we have studied the performance of different normalization techniques and fusion rules in the context of a multimodal biometric system based on the face, fingerprint and hand-geometry traits of a user. Experiments conducted on a database of 100 users indicate that the application of min–max, z-score, and tanh normalization schemes followed by a simple sum of scores fusion method results in better recognition performance compared to other methods. However, experiments also reveal that the min–max and z-score normalization techniques are sensitive to outliers in the data, highlighting the need for a robust and efficient normalization procedure like the tanh normalization. It was also observed that multimodal systems utilizing user-specific weights perform better compared to systems that assign the same set of weights to the multiple biometric traits of all users.
Conference Paper
Biometric recognition refers to an automatic recognition of individuals based on a feature vector (s) derived from their physiological and/or behavioral characteristic. Biometric recognition systems should provide a reliable personal recognition schemes to either confirm or determine the identity of an individual. Applications of such a system include computer systems security, secure electronic banking, mobile phones, credit cards, secure access to buildings, health and social services. By using biometrics a person could be identified based on "who she/he is" rather then "what she/he has" (card, token, key) or "what she/he knows" (password, PIN). In this paper, a brief overview of biometric methods, both unimodal and multimodal, and their advantages and disadvantages, will be presented.