Belen Fernandez-Saavedra

Iris Biometrics for Embedded Systems

In many applications user authentication has to be carried out by portable devices. Usually these devices are personal tokens carried by users, which have many constraints regarding their computational performance, occupied area, and power consumption. These kinds of devices must deal with such constraints, while also maintaining high performance rates in the authentication process. This paper provides solutions to designing such personal tokens where biometric authentication is required. In this paper, iris biometrics have been chosen to be implemented due to the low error rates and the robustness their algorithms provide. Several design alternatives are presented, and their analyses are reported. With these results, most of the needs required for the development of an innovative identification product are covered. Results indicate that the architectures proposed herein are faster (up to 20 times), and are capable of obtaining error rates equivalent to those based on computer solutions. Simultaneously, the security and cost for large quantities are also improved.

Evaluation methodology for analyzing usability factors in biometrics

In recent years, biometrics is more and more used in many security applications. This fact has led suppliers and researches to analyze biometric algorithms power and vulnerabilities, as to improve the feasibility of this technology. Nevertheless, as many authors claim, biometric performance does also depend on other factors such as usability and/or user acceptance, which can influence significantly their performance. Only a few of these factors have already been studied, using specific approaches and only for certain biometric modalities, such us fingerprint and face. However, there is not a general and independent methodology implemented to assess how these factors affect biometric system performance and to produce intercomparable results. Based on previous works and following procedures and requirements addressed in the International Standard ISO/IEC 19795–2 for scenario evaluations [1], authors have developed a general methodology to analyze end-to-end system performance when some usability factors are modified. Such factors cover different ways of presenting biometric characteristics to the sensor and also the biometric characteristic variability caused by illness or climatic changes. A generic and controlled scenario has been modelled to carry on all sets of trials. Then, the methodology has been particularized defining specific protocols, methods and considerations for each parameter to assess. Furthermore, details for analysing these parameters through different modalities have been defined. In addition, this methodology has been checked for one modality considering different usability aspects in order to obtain the feedback that is necessary to test its validity and viability and to detect points of interest for improvement. Results, main conclusions and suggestions for test operators will be presented.

Small fingerprint scanners used in mobile devices: the impact on biometric performance

Biometrics has burst into mobile technology. Fingerprint scanners are being embedded in smartphones and tablets supplying these devices with the security and usability provided by biometric authentication mechanisms. However, performance results obtained by biometric systems cannot be extrapolated to mobile devices. The conditions change, especially at capture process, due to the reduced sensing area of the scanners used. The impact of small fingerprint scanners on the quality and biometric performance of the system is studied. A database using three different fingerprint scanners has been collected and reduced-size images (i.e. 12 × 12 mm 2 , 10 × 10 mm 2 and 8 × 8 mm 2 ) have been modelled by cropping the original ones. Performance testing has been conducted using one public and one commercial algorithm, and considering two application scenarios. One scenario in which enrolment and authentication are executed using the same small sensor included in the mobile device (i.e. cropped image against cropped image) and a second scenario in which enrolment is executed using an external larger sensor and authentication is done using the mobile device sensor (i.e. full image against cropped image). Results show the gradual worsening of quality and error rates as the size of the fingerprint scanner is reduced revealing a significant difference between the application scenarios analysed.

Quality Measurements for Iris Images in Biometrics

A method for analyzing the quality of iris images is presented in this paper. Iris Biometrics for personal identification is based on capturing an eye photograph and obtaining a feature vector that allows to identify a human being. Sometimes these images are not good enough due to a variety of reasons: e.g. out of focus, iris covered by eyelashes or eyelids, etc. It is very important to know if an image has enough quality before applying the feature extraction algorithms, in order to avoid bad results and/or save processing power. The methods designed by the authors, have been evaluated with the images of the ICE 2005 database.

Changes to vascular biometric system security & performance

When considering biometrics for a personal identification system, different modalities can be considered. The final selection will depend on specific application requirements. From the several modalities existing nowadays, vascular systems have appeared on the scene lately. Vein-based identification is claimed to be as reliable as fingerprint or iris identification, but with the usability of hand geometry or even the face. As with any other biometric modalities, vascular solutions have to be deeply analyzed for all relevant factors that could affect their performance or the security level achieved. This will show the analysis of the performance and security achieved by a commercial vascular biometric system, when being applied in several real-world scenarios. These scenarios will cover typical office environments to the more extreme environments, such as extreme light and temperatures, as those suffered in a banking ATM, or with high humidity such as in a gym. The results will show the strengths of this modality, as well as those points where further improvements are needed.

Evaluation methodology for fake samples detection in biometrics

Nowadays biometrics is being used in many applications where security is required. This fact causes that new threatens have appeared and that the number of attempts to break biometric systems has increased. From all potential attacks, those involving damage or thefts to users are the most worrying. Most of them could be avoided if acquisition sensors would have suitable approaches for aliveness detection at the capture process. Many providers claim that their products support these methods but unfortunately it has been discovered that some products do not detect fake samples. In this paper a methodology based on Common Criteria is given to evaluate, in an independent way, whether biometric capture devices implement methods for fake samples detection, and till which extent such methods are effective. This methodology has been tested with sensors from different modalities.

Standardised system for automatic remote evaluation of biometric algorithms

It is difficult to generate significant results when evaluating algorithms developed in biometrics. Additionally, when third parties, such as system integrators, want to compare results among different algorithm providers, they cannot do so easily for several reasons: difficulty accessing large databases, inability to exchange biometric data among researchers due to data protection laws in some countries, or the lack comprehensive and standardised evaluation reports. This paper presents a new performance evaluation system for biometric systems that is secure, automatic and remote. This system has been developed using current standards developed within ISO/IEC JTC1/SC37 for data Formats, Application Program Interfaces (APIs) and evaluation methodology. Standardised technology is able to provide developers in biometrics and third parties with a way to perform comprehensive evaluations remotely and with 24/7 availability without compromising the privacy of the individuals included in the test crew. The solution described here offers the developers the ability to evaluate large databases that are stored in a secured centralised server. As this system is modality-independent, researchers can use the same protocol to perform different evaluations, and therefore lower the overhead costs for testing purposes. Additionally, such protocols can be plugged directly into end-user applications, minimising technology transfer costs. The system is described by block diagrams as well as flowcharts.

Quality metrics influence on iris recognition systems performance

When working with biometrics, regardless of the modality, it is essential to consider data quality, as it can limit the accuracy of recognition systems. In this context, the assessment of biometric samples quality by defining quality metrics can be used to enhance the performance and functionality of a biometric system. In this work, the influence of ISO/IEC 29794–6 quality metrics on iris recognition systems performance has been analysed. To carry out the analysis, the corresponding metrics have been calculated according to the computational method detailed in the standard and a quality score has been obtained. Then, in order to determine the metrics influence, an open source iris recognition system as well as one publicly available dataset, have been used. Results show to what extent an iris recognition system can be affected by input images quality, and also the most influential quality metrics on recognition accuracy.

Usability Evaluation of Fingerprint Based Access Control Systems

Studying usability of biometric products is of great interest for learning in advance how users will behave once the system is installed. This paper presents the work carried out for the evaluation of the usability of five access control systems based on fingerprint recognition. Users are grouped by ages, and a homogeneous representation of ages and sex are considered. Results will show how different devices can present different results considering usability, and conclusions will provide the lessons learned as to be applied in following evaluations.

User-friendly biometric camera for speeding iris recognition systems

Among biometric modalities existing nowadays for person identification, iris pattern recognition has reputed as one of the best performance systems. Unfortunately it has the lack of the user-friendliness that can be seen in hand geometry or face systems. To overcome these problems we propose a new kind of camera to capture iris image.