Informatica 37 (2013) 115–122 115

Building Cloud-based Biometric Services

Peter Peer and Jernej Bule
Faculty of Computer and Information Science
University of Ljubljana, Tržaška cesta 25, SI-1000 Ljubljana, Slovenia
E-mail: {jernej.bule, peter.peer}@fri.uni-lj.si

Jerneja Žganec Gros and Vitomir Štruc1

Alpineon d.o.o., Ulica Iga Grudna 15, SI-1000, Slovenia
1
Faculty of Electrical Engineering, University of Ljubljana, Tržaška cesta 25, SI-1000 Ljubljana, Slovenia
E-mail: {vitomir.struc, jerneja.gros}@alpineon.com

Keywords: biometrics, cloud computing, cloud integration, SaaS, fingerprint recognition

Received: December 4, 2012

Over the next few years the amount of biometric data being at the disposal of various agencies and
authentication service providers is expected to grow significantly. Such quantities of data require not
only enormous amounts of storage but unprecedented processing power as well. To be able to face this
future challenges more and more people are looking towards cloud computing, which can address these
challenges quite effectively with its seemingly unlimited storage capacity, rapid data distribution and
parallel processing capabilities. Since the available literature on how to implement cloud-based
biometric services is extremely scarce, this paper capitalizes on the most important challenges
encountered during the development work on biometric services, presents the most important standards
and recommendations pertaining to biometric services in the cloud and ultimately, elaborates on the
potential value of cloud-based biometric solutions by presenting a few existing (commercial) examples.
In the final part of the paper, a case study on fingerprint recognition in the cloud and its integration into
the e-learning environment Moodle is presented.
Povzetek: Predstavljene so metode za biometrično razpoznavanje oseb, realizirane v oblaku.

1 Introduction
When talking about Internet authentication, in most necessary to devise highly scalable biometric technology,
cases, people are still talking about passwords. One of capable of operating on enormous amounts of data,
the biggest problems with current authentication which, in turn, induces the need for sufficient storage
approaches is the existence of too many password- capacity and significant processing power.
account pairings for each user, which leads to forgetting The first solution that comes to mind with respect to
or using the same username and password for multiple the outlined issues is moving the existing biometric
sites [1]. A possible solution to this problem can be technology to a cloud platform that ensures appropriate
found in the use of biometrics [2]. Biometric scalability of the technology, sufficient amounts of
authentication techniques, which try to validate the storage, parallel processing capabilities, and with the
identity of an user based on his/her physiological or widespread availability of mobile devices also provides
behavioral traits, are already quite widely used for local an accessible entry point for various applications and
authentication purposes (for private use), while their use services that rely on mobile clients. Hence, cloud
on the Internet is still relatively modest. The main reason computing is capable of addressing issues related to the
for this setting is open issues pertaining mainly to the next generation of biometric technology, but at the same
accessibility and scalability of existing biometric time, offers new application possibilities for the existing
technology. generation of biometric systems [4], [5].
Similar issues are also encountered in other However, moving the existing biometric technology
deployment domains of biometric technology, such as to the cloud is a nontrivial task. Developers attempting to
forensics, law-enforcement and alike. For example, tackle this task need to be aware of:
according to [3], the biometric databases of the Federal
Bureau of Investigation, the US State Department,  the most common challenges and obstacles
Department of Defense, or the Department of Homeland encountered, when moving the technology to a
Security are expected to grow significantly over the next cloud platform,
few yours to accommodate several hundred millions (or
even billions) of identities. Such expectations make it
116 Informatica 37 (2013) 115–122 P. Peer et al.

 standards and recommendations pertaining to both addition to presenting a case study on a specific
cloud-based services as well as biometrics in modality.
general, and
 existing solutions that can be analysed for 3 Biometrics and cloud computing
examples of good practices.

This paper tries to elaborate on the above listed issues

3.1 Cloud computing
and provide potential developers with some basic Cloud computing is a computing model, where resources
guidelines on how to move biometric technology to a such as computing power, storage, network and software
cloud platform. It describes the most common pitfalls are abstracted and provided as services on the internet in
encountered in the development work and provides some a remotely accessible fashion [13].
directions for their avoidance. Additionally, it presents a NIST defines five key characteristics of cloud
case study on fingerprint recognition in the cloud, where computing [14]:
the presented guidelines are put into action. The main
motivation for the paper stems from our own work in the  Rapid elasticity - elasticity is defined as the ability
field of cloud-based biometric services1 and the fact that to scale resources both up and down as needed. To
the available literature on this field is extremely limited. the consumer, the cloud appears to be infinite, and
The rest of the paper is structured as follows. In the consumer can purchase as much or as little
Section 2 the existing literature pertaining to biometrics computing as needed [14].
in the cloud is surveyed and differences with this paper  Measured services – certain aspects of the cloud
are highlighted. In Section 3 some basic characteristics of service are controlled and monitored by the cloud
cloud computing, biometrics, and cloud-based biometric provider. This is crucial for billing, access control,
services are presented. In Section 4 issues to consider resource optimization, capacity planning and other
when developing cloud-based biometrics are elaborated tasks [14].
on. In Section 5 a case study on fingerprint recognition in  On-demand self-service - a consumer can use cloud
the cloud is presented and, finally, the paper is concluded services as needed without any human interaction
with some final comments and directions for future work with the cloud provider [14].
in Section 6.
 Ubiquitous network access - the cloud provider’s
capabilities are available over the network and can
2 Related work be accessed by various clients through standard
Cloud computing is a highly active field of research and mechanisms [14].
development, which gained popularity only a few years  Resource pooling - allows a cloud provider to serve
ago. Since the field covers a wide range of areas relating its consumers via a multi-tenant model. Physical
to all levels of cloud computing (i.e. PaaS, IaaS, and and virtual resources are assigned and reassigned
SaaS), it is only natural that not all possible aspects of according to consumer demand. There is a sense of
the field is appropriately covered in the available location independence in that the customer
scientific literature. This is also true for cloud-based generally has no control or knowledge over the
biometrics. exact location of the provided resources, but may be
While there are some papers addressing this topic, able to specify location [14].
they are commonly concerned with specific aspects of
the technology and neglect the bigger picture. The work Clearly, cloud computing has several desirable
of Gonzales et. al [7], for example, addresses cloud- characteristics, which make the cloud platform highly
based biometrics, but focuses on how to protect suitable for various applications, including biometrics.
biometric data from miss-use through a crypto-biometric
system. A similar topic is also discussed by Vallabhu and
Satyanarayana in [8]. Other researchers focus more on
3.2 Biometric systems
developing biometric technology for a certain biometric Biometric recognition systems represent pattern
modality and present cloud computing as a possible use- recognition systems, capable of recognizing individuals
case [9], [10]. This paper, on the other hand, tries to based on their physiological or behavioural traits [2].
cover different aspects of cloud-based biometrics and is These traits are considered to be unique to each
equally interested in legal (e.g., issues relating to data individual and unlike knowledge or token-based security
protection, data retention etc.) as well as technical issues. mechanisms cannot be forgotten, lost or stolen. The most
From this point of view, the topic of the paper is more common traits used for biometric recognition are: faces,
closely related to the work of Senk and Dotzler [11] or fingerprints, irises, palm-prints, speech etc.
Kohlwey et. al [12], where biometrics and cloud
computing are also discussed in a broader context in

1
Conducted in the scope of the KC CLASS (CLoud Assisted ServiceS)
project. [6]
Building Cloud-based Biometric Services Informatica 37 (2013) 115–122 117

Figure 1: Block diagram of a typical biometric recognition system.

Biometric systems typically conduct one of two
tasks: identification or verification/authentication. The 3.3 Biometrics in the cloud
verification/authentication task tries to validate the
As emphasized in the previous section, there are certain
identity claim of the user currently presented to the
aspects of biometric systems that are specific to cloud
system, while the identification task tries to determine,
computing. First of all, the biometric engine2 is located
which of the registered user the acquired “live” biometric
in the cloud and not on some local processing unit, as it
sample corresponds to. Hence, the identification problem
is the case with traditional (e.g. access control) biometric
is commonly considered to be a one-to-N matching
recognition systems. This characteristic makes the cloud-
problem, while the verification/authentication problem is
based biometric technology broadly accessible and
considered to be a one-to-one matching problem.
provides the necessary means for integration in other
Biometric systems always comprise the same basic
security and/or consumer applications. Second of all,
components regardless of whether they are designed for
storing biometric data in the cloud makes the system
the cloud or any other platform. These components,
highly scalable and allows quick and reliable adaptation
which are also shown in Fig. 1 for the case of a face
of the technology to an increasing user base [3].
recognition system, include [2], [4]:
On the other hand, storing biometric data in the
i) a data acquisition component (or sensor) that cloud may raise privacy concerns and may not be in
captures a still image or video sequence of a user accordance with national legislation. Last but not least, a
trying either to enrol into the system or to use the cloud implementation of biometric technology may
system for authentication/identification purposes, harvest all merits of the cloud, such as real-time and
parallel processing capabilities, billing by usage etc. [3].
ii) a template generation component that uses machine
All of the presented characteristics make cloud-based
learning, computer vision and pattern recognition
biometric recognition technology extremely appealing.
techniques to derive a biometric template from the
When developing biometric technology for the
input data,
cloud, one needs to make a number of design choices.
iii) a database of biometric templates belonging to Probably the most important choice is, which
enrolled/registered users, and components to move to the cloud and which to
implement locally. A review of some existing market
iv) a matching component that compares the biometric solutions ([15], [16], [17], [18], [19]) from the field of
template derived from the “live” image with the cloud-based biometrics reveals that most often both the
appropriate template(s) stored in the database of the biometric engine as well as the biometric database is
system and based on the outcome makes a decision moved to the cloud. The commercial solutions typically
regarding the identity of the user currently operate on the principle of the client-server model. The
presented to the system. local client (e.g. on the user’s computer) is responsible
for capturing a biometric sample of the user and sending
While the basic layout of a biometric recognition system it to the server (hosted in the cloud), where the matching
is more or less the same on any platform (and biometric process is executed. For the safety of the network traffic
modality), there are, however, a number of aspects that between the client and the server designated security
are specific to the cloud. These aspects will be discussed protocols are commonly used.
in more detail in the next section.

2
We will refer to the template generation and matching
components as the biometric engine in the remainder of the paper.
118 Informatica 37 (2013) 115–122 P. Peer et al.

While the presented configuration makes full use of 4.2 Standards and recommendations
the merits of the cloud platform, it may not be
There are several standards and recommendations that
conformant with the local legislation. Therefore, the
are relevant in the context of both biometric recognition
possibility of using a locally hosted database needs to be
as well as cloud computing. These include internet
considered when designing a cloud-based biometric
protocols, data formats, communication and security
system. Such a setting may limit the scalability of the
protocols, recommendations for cloud application design,
technology to a certain extent, but is reasonable as it
recommendations for biometric technology design etc.
makes potential market-ready technology more easily
Since this field is too broad to be covered completely, the
adjustable to currently existing legislation. Another
focus of this paper is only on a small number of
possible solution to the legislation problem could also be
important standards related to biometric recognition
found in the use hybrid clouds.
technology in the cloud.
The first group of standards of interest for every
4 Integrating biometrics in the cloud developer working in the field of biometric recognition
are standards that allow for interoperability among
4.1 Challenges and obstacles different vendors (e.g. [22], [23]). These standards define
interchange formats for biometric data and (next to
When developing biometric technology for the cloud,
interoperability) also enable consolidation of different
one inevitably encounters a number of challenges and
biometric databases. The standard in [23], for example,
obstacles that need to be addressed. Next to meeting
specifies interchange formats for face images and as such
performance criteria and selecting the most suitable
defines full-frontal and token face images (defined by the
platform for the development work, current legislation
location of the eyes) and ensures that enrolled images
pertaining to cloud computing and biometrics in general,
meet a sufficient quality standard for arbitrary face
privacy concerns and data protection issues all represent
recognition technology. Similar standards also exist for
major challenges for the development process [4].
other biometric traits [24].
The challenges pointed out above are addressed in
The second group of standards of relevance to cloud-
different ways. The performance of the biometric
based biometrics is the OASIS standard for Biometric
recognition technology can systematically be evaluated
Identity Assurance Services (BIAS) [25]. The open
using established reproducible scientific methodology.
standard defines all specifications for SOAP-based
Here, publicly available databases with predefined
biometric services and is conveniently supported by a
experimental protocols and performance criteria are
reference implementation (for fingerprints) provided by
typically employed to produce performance estimates
NIST. The ISO/IEC JTC 001/SC 37 has just recently
that can be compared with performance estimates of
approved a project to internationalize the above
previously assessed technology.
mentioned BIAS standard.
The platform used in the development work is
commonly selected according to ones preferences or with
respect to the planned characteristics of the final product 4.3 Deployment possibilities and existing
(i.e. deployable in a private or public cloud etc.). solutions
When it comes to legal, privacy and data protection Cloud-based biometric technology offers attractive
concerns, there are usually no universal solutions, as they deployment possibilities, such as smart spaces, ambient
differ from country to country. In the case of Slovenia, intelligence environments, access control applications,
for example, the information officer has composed mobile application, and alike. While traditional (locally
several guidelines/recommendations both for the cloud as deployed) technology has been around for some time
well as biometric technology. The recommendations now, cloud-based biometric recognition technology is
relating to biometric technology, biometric data relatively new. There are, however, a number of existing
protection and template storage can be found in [20] and solutions already on the market, these include (among
fall in the domain of ZVOP-1 (in Slovenian: Zakon o others) the solutions by Animetrics [15], BioID [16] and,
varstvu osebnih podatkov), while the guidelines for cloud of course, Face.com [17], which has recently been
computing are accessible from [21]. acquired by Facebook.

Figure 2: Simplified block diagram of biometric registration and verification.

Building Cloud-based Biometric Services Informatica 37 (2013) 115–122 119

5 A case study: fingerprint implemented locally. For our case study, we decided to
move the biometric engine as well as the biometric
recognition in the cloud database to the cloud. A block diagram of the complete
cloud-based biometric service design is shown in Fig. 3.
5.1 Goal and setup Note that the verification process with the described
The goal of the case study presented in the remainder is design is conducted using the following scenario:
to put the general guidelines presented in the previous i) the fingerprint of a given user is first captured via a
sections into practice and provide more detailed fingerprint scanner (here scanner libraries that allow
(technical) information on the process of integrating capturing fingerprint images need to be integrated
biometric technology into a cloud platform. The basis of into the local (desktop or/and web) application);
the case study represents a prototype fingerprint ii) the application then communicates through a
recognition systems, named FingerIdent [26]. A local test (REST) API with the biometric web service hosted
version of this prototype system is already installed at the in the cloud and sends an encoded image to the
Faculty of Computer and Information Science, fingerprint processing library (i.e. FingerIdent
University of Ljubljana, in front of the Computer Vision library) that provides the functionality for the cloud
Laboratory. service;
The functionality of the existing local version of the iii) the transmitted fingerprint image is processed in the
FingerIdent system can be divided into two main cloud and finally the result is sent back to the local
categories: application.
i) user registration (enrollment), during which a The security of the presented solution is provided on
biometric template of a given user is constructed different levels through:
and stored in the system’s database, and
 the use of the HTTPS protocol for data transfer,
ii) user verification, during which the identity claim of
 the use of certificates (the SSL protocol),
a given user is validated.
 the encryption of passwords and other data (such as
The registration process uses a fingerprint reader to biometric templates) in the database, and
capture the (biometric) fingerprint data. In the next phase  the protection of the access to the cloud-service
the quality of the captured sample is evaluated and if it is with a complex 40-digit password.
found to be adequate, the system extracts features from it
and stores them in the form of a biometric template in the The cloud-based service is designed modularly,
database. During the verification process features from which makes upgrading the service a relatively simple
the captured “live” fingerprint are again extracted and task. Equally important is the fact that the same design is
compared to those stored in the database. The also suitable for other biometric modalities and allows
comparison is made based on pattern matching for devising multi-modal person authentication as well.
procedures, which form the foundation for the validation
of the identity claim. An illustration of both functions is 5.3 Moodle with fingerprint verification
shown in Fig. 2. To demonstrate the effectiveness of the presented
To reach the goal of devising a cloud-based solution and to provide a proof-of-concept, the e-learning
biometric service, one needs to migrate the presented environment Moodle [27] is augmented with biometric
functionality of the local FingerIdent system to the cloud authentication capabilities by integrating it with the
and provide the necessary infrastructure for accessing the cloud-based fingerprint verification service.
biometric service. Details on this procedure are given in Since Moodle is also designed modularly, the
the next section. biometric authentication procedure is implemented as an
additional (optional) authentication scheme, which can
5.2 Designing cloud biometric services complement the existing procedures and provide an
It was emphasized in Section 3.3 that a decision has to be additional level of access security. A block diagram of
made with respect to which components of the biometric the integration is shown in Fig. 4.
system should be moved to the cloud and which
120 Informatica 37 (2013) 115–122 P. Peer et al.

Figure 3: Scheme of the biometric verification system in the cloud.

Figure 4: Cloud fingerprint verification in Moodle.

The main problem faced during integration is the The solution developed in the scope of this case
compatibility of various fingerprint readers with different study uses an ActiveX component to access the
browsers. Each manufacturer of fingerprint readers offers hardware. ActiveX components are officially supported
their own protocols and libraries to access the only on Internet Explorer, which represents a weakness
corresponding hardware. A standard is not yet available. in the implementation. As future work, an extension of
the presented solution is planned, so it can work with
Building Cloud-based
based Biometric Services Informatica 37 (2013) 115–122 121

other popular browsers, such as Firefox, Opera or References

Chrome too.
After the integration of the fingerprint authentication [1] D. Balfanz et al., "The future of authentication",
service into the Moodle framework, the Moodle login IEEE Security & Privacy
Privacy, vol. 10, pp. 22-27, 2012.
screen was modified to account for the t added [2] A.K. Jain, A. Ross, and S. Prabhakar, “An
functionality. The result of this procedure is shown in Introduction to Biometric Recognition,” IEEE
Fig. 5. Note
ote how the added biometric authentication Transactions on Circuits and Video Technology
Technology,
functionality seamlessly integrates into the existing vol. 14, no. 1, pp. 4-20,
20, 2004.
framework. [3] E. Kohlwey, A. Sussman, J. Trost, and A. Maurer,
“Leveraging the Cloud for Big Data Biometrics:
Meeting the performance requirements of the Next
Generation Biometric Systems,” in Proceeding of
the IEEE World Congress on Services
Services, pp. 597-601,
2011.
[4] V. Štruc and J. Žganec
Žganec-Gros, “Developing Face
Recognition Technology for the KC Class
Biometrics service,” in: CLASS Conference 2012,
pp. 68-75, 2012.
[5] J. Bule and P. Peer, “Fingerprint Verification as a
Service in KC CLASS,” in: CLASS Conference
2012, pp. 76-82, 2012.
[6] The KC Class project project, available from:
http://www.kc-class.eu/,, last visited: 5.12.2012.
[7] D. Gonzales Martinez, F.J. Gonzels Castano, E.
Argones Rua, J.L. Ala Castro, D.A. Rodriguez
Silva, “Secure Crypto-Biometric
Biometric System for Cloud
Computing,” in: International Workshop on
Securing Services on the Cloud
Cloud, pp. 38-45, 2011.
[8] H. Vallabhu and R.V. Satyanarayana, “Biometric
Authentication as a Service on Cloud: Novel
Solution,” International JoJournal of Soft Computing
and Engineering,, vol. 2, no. 4, pp. 163
163-165, 2012.
[9] S. Suryadevara, S. Kapoor, S. Dhatterwal, R. Naaz
Figure 5: Customized Moodle login.
login and A. Sharma, “Tongue ue as a Biometric Visualizes
New Prospects of Cloud Computing Security,” in:
International Conference on Inform Information and
6 Conclusion Network Technology,, vol. 4, 2011.
Cloud based biometric services have an enormous [10] S.N.S. Raghava,, “Iris Recognition on Handoop: a
potential market value and as such attract the interest of Biometrics System Implementation on Cloud
research and development groups from all around the Computing,” in: Proceedings of IEEE CCIS, 2011.
world. In this paper some directions on how to move [11] C. Senk and F. Dotzler, “Biometric Authentication
existing biometric technology to a cloud platform
plat were as a Service for Enterprise
terprise Identity Management
presented. Issues that need to be considered when Deployment: A Data Protection Perspective,” in:
designing cloud-based
based biometric services have been International Conference on Availability,
presented and a case study, where a cloud-basedcloud Reliability and Security,, pp. 43-50, 2011.
fingerprint service was developed and integrated with the [12] E. Kohlwey, A. Sussman, J. Trost Trost, and A. Maurer,
e-learning
learning framework Moodle was described
describ as well. As “Leveraging
Leveraging the Cloud for Big Data BiometricsBiometrics:
part of our future work we plan to migrate more Meeting the Performance Requirements of the Next
biometric modalities to the cloud and, if possible, devise Generation Biometric Systems
Systems,” in: IEEE World
a multi-modal cloud-based
based biometric solution Congres on Services, pp. 597597-601, 2011.
[13] D.M. Dakhane and A.A. Arokar, “Data Secuity in
Acknowledgements Cloud Computing for Biometric Application,”
International Journal of Sci Scientific & Engineering
The work presented in this paper was supported by the Research,, vol. 3, no. 6, pp. 11-4, 2012.
European Union, European
ropean Regional Fund, within the [14] Cloud computing use case discussion group, “Cloud
scope of the framework of the Operational Programme Computing Use Cases: White Paper” available
for Strengthening Regional Development Potentials
Pote for from: http://cloudusecases.org/
http://cloudusecases.org/, last visited:
the Period 2007-2013 contract No. 3211-10-000467
3211 (KC 05.12.2012.
Class), the postdoctoral
ral project BAMBI with ARRS ID [15] Homepage
omepage of the Animetrics cloud cloud-based face
Z2-4214. recognition solution
solution, available from:
122 Informatica 37 (2013) 115–122 P. Peer et al.

http://animetrics.com/cloud-face-recognition-
services/, last visited: 03.10.2012.
[16] Homepage of the BioID cloud-based biometric
recognition solution, available from:
http://www.bioid.com/, last visited: 03.10.2012.
[17] Homepage of the Face.com cloud-based face
recognition solution, available from:
http://face.com/, last visited: 03.10.2012.
[18] Homepage of Ceelox ID Online, available from:
http://www.ceelox.com/ceeloxidonline.html, last
visited: 05.12.2012.
[19] Homepage of PasswordBank IDaaS, available from:
http://www.passwordbank.com/passwordbank-
private-cloud, last visited: 05.12.2012.
[20] Homepage of the Slovenian Information
Commissioner, biometrics, available from:
https://www.ip-rs.si/varstvo-osebnih-
podatkov/informacijske-tehnologije-in-osebni-
podatki/biometrija/, last visited: 03.10.2012.
[21] Information Commissioner, Cloud Security
Alliance Slovenia Chapter, Slovenski institut za
revizijo, Slovenski odsek ISACA, Zavod e-Oblak,
Eurocloud Slovenia, “Varstvo osebnih podatkov &
računalništvo v oblaku,” pp. 31, 2012, available
from: https://www.ip-rs.si/fileadmin/user_upload/
Pdf/smernice/Smernice_rac_v_oblaku.pdf, last
visited: 03.10.2012.
[22] Information technology, “Biometric data
interchange formats – Part 5: Face image analysis,”
Documents ISO/IEC 19794-5:2005, 2004, available
from: http://www.iso.org, last visited: 03.10.2012.
[23] Information technology, “Face recognition format
for data interchange,” Document 385-2004 ANSI
INCITS, 2004, available from: http://www.iso.org,
last visited: 03.10.2012.
[24] NIST standard, ANSI/NIST-ITL 1-2011, NIST
Special Publication 500-290, Data Format
for the Interchange of Fingerprint, Facial & Other
Biometric Information, available from:
http://www.nist.gov/itl/iad/ig/ansi_standard.cfm,
last visited: 03.10.2012.
[25] OASIS standard, “Biometric Identity Assurance
Services (BIAS) SOAP Profile Version 1.0,” pp.
210, May 2012, available from: https://www.oasis-
open.org/committees/tc_home.php?wg_abbrev=bia
s, last visited: 03.10.2012.
[26] M. Tovšak, J. Bule, P. Peer, "Upgrading a system
for verification based on fingerprints," in:
Electrotechnical and Computer Science Conference
(ERK), vol. B, pp. 135-138, 2011.
[27] Moodle, open-source e-learning software platform,
available from: http://moodle.org, last visited:
06.12.2012.