Rosario Arjona

A hardware solution for real-time intelligent fingerprint acquisition

The first step in any fingerprint recognition system is the fingerprint acquisition. A well-acquired fingerprint image results in high-resolution accuracy and low computational effort of processing. Hence, it is very useful for the recognition system to evaluate recognition confidence level to request new fingerprint samples if the confidence level is low, and to facilitate recognition process if the confidence level is high. This paper presents a hardware solution to ensure a successful and friendly acquisition of the fingerprint image, which can be incorporated at low cost into an embedded fingerprint recognition system due to its small size and high speed. The solution implements a novel technique based on directional image processing that allows not only the estimation of fingerprint image quality, but also the extraction of useful information (in particular, singular points). The digital architecture of the module is detailed and their features in terms of resource consumption and processing speed are illustrated with implementation results into FPGAs from Xilinx. Performance of the solution has been verified with fingerprints from several standard databases that have been acquired with sensors of different sizes and technologies (optical, capacitive, and thermal sweeping).

A digital circuit for extracting singular points from fingerprint images

Since singular point extraction plays an important role in many fingerprint recognition systems, a digital circuit to implement such processing is presented herein. A novel algorithm that combines hardware efficiency with precision in the extraction of the points has been developed. The circuit architecture contains three main building blocks to carry out the three main stages of the algorithm: extraction of a partitioned directional image, smoothing, and searching for the patterns associated with singular points. The circuit processes the pixels in a serial way, following a pipeline scheme and executing in parallel several operations. The design flow employed has been supported by CAD tools. It starts with high-level descriptions and ends with the hardware prototyping into a FPGA from Xilinx.

Model-based design for selecting fingerprint recognition algorithms for embedded systems

Most of contributions for biometric recognition solutions (and specifically for fingerprint recognition) are implemented in software on PC or similar platforms. However, the wide spread of embedded systems means that fingerprint embedded systems will be progressively demanded and, hence, hardware dedicated solutions are needed to satisfy their constraints. CAD tools from Matlab-Simulink ease hardware design for embedded systems because automatize the design process from high-level descriptions to device implementation. Verification of results is set at different abstraction levels (high-level description, hardware code simulation, and device implementation). This paper shows how a design flow based on models facilitates the selection of algorithms for fingerprint embedded systems. In particular, the search of a solution for directional image extraction suitable for its application to singular point extraction is detailed. Implementation results in terms of area occupation and timing are presented for different Xilinx FPGAs.

Fuzzy models for fingerprint description

Fuzzy models, traditionally used in the control field to model controllers or plants behavior, are used in this work to describe fingerprint images. The textures, in this case the directions of the fingerprint ridges, are described for the whole image by fuzzy if-then rules whose antecedents consider a part of the image and the consequent is the associated dominant texture. This low-level fuzzy model allows extracting higher-level information about the fingerprint, such as the existence of singular points and their fuzzy position within the image. This is exploited in two applications: to provide comprehensive information for users of unattended automatic recognition systems and to extract linguistic patterns to classify fingerprints.

A PUF- and Biometric-Based Lightweight Hardware Solution to Increase Security at Sensor Nodes

Security is essential in sensor nodes which acquire and transmit sensitive data. However, the constraints of processing, memory and power consumption are very high in these nodes. Cryptographic algorithms based on symmetric key are very suitable for them. The drawback is that secure storage of secret keys is required. In this work, a low-cost solution is presented to obfuscate secret keys with Physically Unclonable Functions (PUFs), which exploit the hardware identity of the node. In addition, a lightweight fingerprint recognition solution is proposed, which can be implemented in low-cost sensor nodes. Since biometric data of individuals are sensitive, they are also obfuscated with PUFs. Both solutions allow authenticating the origin of the sensed data with a proposed dual-factor authentication protocol. One factor is the unique physical identity of the trusted sensor node that measures them. The other factor is the physical presence of the legitimate individual in charge of authorizing their transmission. Experimental results are included to prove how the proposed PUF-based solution can be implemented with the SRAMs of commercial Bluetooth Low Energy (BLE) chips which belong to the communication module of the sensor node. Implementation results show how the proposed fingerprint recognition based on the novel texture-based feature named QFingerMap16 (QFM) can be implemented fully inside a low-cost sensor node. Robustness, security and privacy issues at the proposed sensor nodes are discussed and analyzed with experimental results from PUFs and fingerprints taken from public and standard databases.

A dual-factor access control system based on device and user intrinsic identifiers

This paper proposes an access control system based on the simultaneous authentication of what the user has and who the user is. At enrollment phase, the wearable access device (a smart card, key fob, etc.) stores a template that results from the fusion of the intrinsic device identifier and the user biometric identifier. At verification phase, both the device and user identifiers are extracted and matched with the stored template. The device identifier is generated from the start-up values of the SRAM in the device hardware, which are exploited as a Physically Unclonable Function (PUF). Hence, if the device hardware is cloned, the authentic identifier is not generated. The user identifier is obtained from level-1 fingerprint features (directional image and singular points), which are extracted from the fingerprint images captured by the sensor in the access device. Hence, only genuine users with genuine devices are authorized to access and no sensitive information is stored or travels outside the access device. The proposal has been validated by using 560 fingerprints acquired in live by an optical sensor and 560 SRAM-based identifiers.

Dedicated hardware IP module for fingerprint recognition

This work presents a dedicated hardware IP module for fingerprints recognition based on a feature, named QFingerMap, which is very suitable for VLSI design. FPGA implementation results of the IP module are given. A demonstrator has been developed to evaluate the IP module behavior in a real scenario.

A fingerprint biometric cryptosystem in FPGA

This paper presents the implementation of a complete fingerprint biometric cryptosystem in a Field Programmable Gate Array (FPGA). This is possible thanks to the use of a novel fingerprint feature, named QFingerMap, which is binary, length-fixed, and ordered. Security of Authentication on FPGA is further improved because information stored is protected due to the design of a cryptosystem based on Fuzzy Commitment. Several samples of fingers as well as passwords can be fused at feature level with codewords of an error correcting code to generate non-sensitive data. System performance is illustrated with experimental results corresponding to 560 fingerprints acquired in live by an optical sensor and processed by the system in a Xilinx Virtex 6 FPGA. Depending on the realization, more or less accuracy is obtained, being possible a perfect authentication (zero Equal Error Rate), with the advantages of real-time operation, low power consumption, and a very small device.

A Fingerprint Retrieval Technique Using Fuzzy Logic-Based Rules

This paper proposes a global fingerprint feature named QFingerMap that provides fuzzy information about a fingerprint image. A fuzzy rule that combines information from several QFingerMaps is employed to register an individual in a database. Error and penetration rates of a fuzzy retrieval system based on those rules are similar to other systems reported in the literature that are also based on global features. However, the proposed system can be implemented in hardware platforms of very much lower computational resources, offering even lower processing time.

Microelectronics implementation of directional image-based fuzzy templates for fingerprints

Fingerprint orientation image, also called directional image, is a widely used method in fingerprint recognition. It helps in classification (accelerating fingerprint identification process) as well as in preprocessing or processing steps (such as fingerprint enhancement or minutiae extraction). Hence, efficient storage of directional image-based information is relevant to achieve low-cost templates not only for “match on card” but also for “authentication on card” solutions. This paper describes how to obtain a fuzzy model to describe the directional image of a fingerprint and how this model can be implemented in hardware efficiently. The CAD tools of the Xfuzzy 3 environment have been employed to accelerate the fuzzy modeling process as well as to implement the directional image-based template into both an FPGA from Xilinx and an ASIC.