The use of biometrics is spreading into many areas, including financial transactions, consumer electronics, medical devices, industrial and business security, border protection, law enforcement, and military electronics. Rising instances of identify theft, fraud, terrorism, and an increasing emphasis on safety and security are all acting as driving forces.
According to the figures released by Acuity Market Intelligence in its recent market research report, “The Future of Biometrics,” biometrics technology is poised for sustained growth. Annual market revenues are expected to reach $11 billion by 2017, representing a compound annual growth rate (CAGR) of 19.69%.
Frost & Sullivan reports that biometrics in Europe, the Middle East, and Africa earned €216.1 million in 2008 (about $300 million), and the researchers expect the market to reach €1.058 billion by 2015 (about $1.5 billion). This represents a CAGR of 25.5%. Frost & Sullivan also forecasts North American revenue in biometrics to reach $9.44 billion in 2014, up from $2.6 billion in 2007.
Government projects worldwide such as e-passports and national identity schemes spurred on by legislative initiatives will be a major factor in this growth. For the biometrics industry, the emphasis continues to be on practical and low-cost standardized biometric approaches that ensure optimal security, non-invasiveness, and user friendliness.
Several modalities are employed for biometrics for personal identification and verification, each providing different levels of authenticity. These include fingerprinting, facial feature recognition, iris recognition, hand recognition, palm and vein recognition, voice recognition, and signature recognition. More recently, holographic methods have entered the fray. When combined with personal data stored on databases, any of these modalities can greatly improve identification and authentication.
Improved Fingerprint Identification
By far, fingerprinting is the most common biometric method, followed by iris recognition. Fingerprinting is improving on several fronts thanks to better capacitive, thermal, and optical sensing technologies.
For example, the National Institute of Standards and Technology (NIST) has developed an emerging software technology called Automated Feature Extraction and Matching (AFEM) to speed up the manual portion of latent fingerprint identification.
Normally, a fingerprint examiner must first carefully mark the distinguishing features of a full or partial latent print, beginning with the positions where ridges end or branch. After that, the print is entered into the FBI’s Integrated Automated Fingerprint Identification system (IAFIS) for further identification.
The AFEM system was tried last year on a data set of 835 latent prints and 100,000 fingerprints that have been used in real case examinations, and it received high marks for accuracy. Half of the prototypes were accurate at least 80% of the time.
For each fingerprint point, the software extracted the distinguishing features of the latent prints and then compared them against 100,000 fingerprints. For each print, the software provided a list of 50 candidates that the fingerprint specialists compared by hand. Most identities were found within the top 10.
Meanwhile, NIST researchers Mary Theofanos, Brian Stanton, Yee-Yin Choong and Ross Michaels are working closely with the FBI and other law-enforcement agencies to develop a mobile ID platform that eliminates the need for law-enforcement officials to take people back to headquarters for fingerprinting.
With its 2- by 3-in. touchscreen, the prototype Quick Capture platform includes a fingerprint sensor that can take pictures of fingerprints or even faces. It then sends these pictures wirelessly to a central hub for analysis with a minimum number of keypad strokes (Fig. 1).
Continue to next page
BIO-Key International recently announced the successful deployment of an FBI-compliant advanced fingerprint-based authentication solution for the Oklahoma County Sheriff’s Office. It complies with federal security requirements for advanced user authentication from in-vehicle laptop computers accessing FBI Criminal Justice Information Services (CJIS) databases while streamlining the log-on process for law-enforcement officials statewide.
BIO-Key’s Vector Segment Technology (VST) patented fingerprint matching algorithm used to identify the agency offices in this application works in concert with the InterAct Mobile data solution from BIO-Key’s partner Interact Public Safety Systems currently used by Oklahoma law-enforcement officials.
The AES2660 192-Mpixel-wide smart sensor from Authentec complements the company’s AES1660 128-Mpixel-wide sensor for PCs, enabling PC OEMs to choose the appropriate sensor for of each individual product (Fig. 2).
Key features of the AES2660 include integrated LED drivers for multiple LED packaging options, Authentec’s TruNav cursor and menu navigation, multiple battery-friendly operating modes, and support for USB 2.0 and serial peripheral interface (SPI) flash memory interfaces. The chip also includes a current limiter, a voltage regulator, and a clock generator.
Sonovation’s SonicSlide STS3000 is the industry’s thinnest, most durable fingerprint sensor, according to the company. This highly accurate device can handle 10 million swipes (Fig. 3). It integrates all of its components into a single 35- by 14.5- by 0.25-mm module with a single sensing element that measures just 3 by 14 by 0.1 mm.
The STS3000 comprises an array of ceramic piezoelectric transducers and advanced polymers combined in a silicon ASIC. The transducer relies on ultrasound technology for operation. It is formed into micro-sized pillars that mechanically oscillate when an electric field is applied. The oscillations then register in 256 shades of gray to form the ridges and valleys of a fingerprint image.
Some fingerprint sensor manufacturers use a patented light-emitting sensor (LES) technology for high-resolution image generation and high levels of authentication (Fig. 4). Integrated Biometrics says that its LES solution is the only technology to achieve the National Biometric Security Project’s Qualified Products List certification. It features low maintenance, high durability, ease of use, high image quality, and greater performance over time compared to other optical, silicon, and silicon swipe technologies.
Templates are captured at 650 dots/in. and then combined with proprietary image-processing algorithms to provide usable templates where other sensor types fail due to foreign matter on the fingerprint such as oil and grease, skin dryness, and the age of the person being fingerprinted. The LES scanner has been mechanically tested to 1.5 million touches with no visible degradation in image quality.
Notebooks And Laptops
Many new fingerprint sensors are being developed specifically for the portable notebook and laptop computer market, where ease of use, small size, low cost, and durability are key concerns. For example, the all-flat ES603-EF from Egis Technology features a flawless design and surface smoothness for easy image capture. The 192-pixel sensor is just 12 by 1.8 mm in thicknesses ranging from 0.5 to 1.0 mm.
Validity Sensor Inc. says that its VFS301 sensor is used in most notebooks and laptops. Unlike most other sensor types, it is made of a durable plastic and is completely de-coupled from the sensor’s drive circuitry. This allows the sensor to provide large and high-quality print images without compromising OEM customers’ cost targets.
Sensor support circuitry is also improving in performance. STMicroelectronics has demonstrated the first microcontroller IC that provides formal certification for secure document scanning of ID cards and biometric passports. The ST23YR80 secure dual microcontroller has received formal EAL6+ certification, which is the highest security standard applicable to microcontrollers. Based on Common Criteria 3.1 methodology, the certification supports all e-passport norms, including basic access control (BAC) and extended access control (EAC) using the highest crypto-key length.
Continue to next page
The ST23YR80 allows a complete and secure passport transaction in less than 3 seconds using secure embedded software from Sagem Securité called IDeal Pass, which STMicroelectronics calls the fastest competitive solution certified to the EAL5+ standard. The chip includes 390 kbytes of user ROM and 8 kbytes of RAM. In addition, the chip has been certified to conform to the ISO/IEC 14443 standard.
Measuring less than 1 in.2, Oki Semiconductor’s highly integrated ML67Q5250 IC (Fig. 5) is one of the smallest fingerprint sensor support chips available. It is part of Oki’s MK67Q5250 module, which includes a microcontroller unit, memory, fingerprint authentication algorithms, and interface circuitry. The module is used in Authentec’s AES2510 fingerprint imaging sensor to perform fingerprint processing and matching, store biometric control code, application code, and templates, and communicate with a host processor.
According to Alec Melnick, senior product marketing for Oki, “The MK67Q5250 allows a system designer to use this pre-assembled module to quickly evaluate and integrate a system. Factory programming of the MK67Q5250 eliminates the need for writing fingerprinting processing code while maintaining a high level of control over the fingerprint enrollment and verification process.”
Biometrics is also getting into other consumer items besides computers as evidenced by a novel means of authentication. The patented Verif-EYE bio-optical interactive visual sensor from Smart Holograms allows consumers to “self-validate” their foods, cosmetics, and pharmaceutical purchases before they consume them or use them, as such over-the-counter products are vulnerable to counterfeiting and tampering.
Ted Wlazlowski, managing director of Smart Holograms, says that “although significant progress has been made in the packaging of consumer products, there is not a solution for the consumer to confirm product authenticity and safety. The Verif-EYE visual sensor gives consumers the long overdue ability to make sure products are legitimate and safe prior to consumption or use.”
Verif-EYE works by showing a visual holographic image or color that transforms into a different image or color once it detects human breath or a drop of water, depending on how the sensor is programmed. Holographic images or messages can be unique to each product. And according to the company, the technology is virtually impossible to copy in volume because it requires many proprietary materials, processes, and pieces of equipment that are not commercially available.
The Eyes Have It
For highly secure access control and border-crossing biometric identification, iris recognition has proven its mettle as one of the most accurate and foolproof methods for verification. According to a NIST report, iris recognition constitutes a powerful biometric that maintains its accuracy and interoperability with compact images. The report affirms the technology’s potential for large-scale identity management applications such as the U.S. Federal Personal Identification Verification program, cyber security, and counterterrorism.
Iris recognition, after fingerprinting, has emerged as the second most widely supported biometric approach. It rests in large part on the ability of recognition algorithms to process standard images from many available cameras. Images should be captured in a standard format and prepared so they are compact enough for use in a smart card and in transmission across global networks.
NIST established the Iris Exchange (iREX) program in collaboration with the biometric industry to encourage the development of iris recognition algorithms operating on images conforming to the ISO/IEC 19794-6 standard. The first iREX project, iREX I, provides quantitative support for the standard by NIST conducting the largest independently administered test of iris recognition technology to date. The standard specifies the JPEG2000 format as superior in image quality to the conventional JPEG format (Fig. 6). Work is underway at NIST on an iREX II version for improved iris recognition accuracy.
Continue to next page
Sarnoff Corp. has enhanced its iris ID system with the introduction of its next-generation iris-on-the-move (IOM) PassPort walk-through system for high-speed, accurate, and quick image capture for people in motion and from a distance up to 10 feet. Unlike other iris recognition systems that require individuals to stop and look into a scanner, the IOM PassPort patented technology can verify individuals’ identities in milliseconds while they are passing through a portal at up to 30 people per minute.
Researchers also believe that the human face has potential as a good biometric tool for identifying individuals. For example, researchers Adrian Evans and Adrian Moorehouse at London’s Bath University believe that scanning a person’s nose can be an effective way of identification. They say that facial expressions can change features like eyes and ears, whereas nose shapes do not change as much.
The researchers used a photographic system called PhotoFace, developed at the University of the West of England in Bristol and the Imperial College of London, to scan the 3D shapes of volunteers’ noses. Computer software was employed to analyze the shapes according to six main nose shapes: Roman, Greek, Nubian, Hawk, Snub and Turn-up. Instead of using the whole shape of the nose, they used three characteristics in their analysis: the ridge profile, the nose tip, and the nasion of section between the eyes and the tip of the nose (Fig. 7).
Researchers Steven Dakin of the University College of London’s Institute of Opthalmology and Roger Watt of the University of Stirling believe that human faces have natural “barcodes” that can be used as a biometric tool. In a study they performed on the faces of celebrities, they concluded that nearly all of the information we need is contained in horizontal lines, such as the line of the eyebrows, the eyes, and the lips. Further analysis revealed that these features could be simplified into black and white information lines, or, in other words, barcodes.
“Exposed skin on our forehead and cheeks tends to be shiny while the shadows cast in the eye sockets and under the nose tend to be darker,” says Dakin. “The resulting horizontal stripes are reminiscent of supermarket barcodes.”
The Animetrics90 product family from Animetrics Inc. is based on the company’s core algorithms for patent-pending 2D-to-3D face creation technologies, allowing for pose-invariant and lighting-invariant face recognition across a broad range of operating conditions. Animetrics is driving the acceptance of facial recognition biometrics in government, homeland security, and law-enforcement applications.
The company’s technology has been integrated into the iDLMax, a handheld identity management device from MaxID Corp. This product is the first commercially available mobile computing device to incorporate Animetrics’ biometric technology. “Animetrics’ 2D-to-3D face analysis is superior to other face detection and matching algorithms,” says Stephen Grist, MaxID’s CEO.
Software Development Kits Simplify Designs
Software development kits for OEMs developing biometric platforms are available to simplify and expedite a platform’s design. For example, Lithuania-based Neurotechnology offers the VeriLook 4.0 software development kit for biometric facial recognition, as well as the MegaMatcher 3.1 software development kit for large-scale automated fingerprint identification systems. In addition, the company has added facial, palm-print, and iris-recognition capabilities to the MegaMatcher tool suite.
Luxand Inc. also offers its FaceCrop software development kit for automatic face detection and the production of passport-size images with a single function call. It produces avatars or passport images of human faces out of digital snapshots.
And thanks to sketching software developed by Christopher Solomon at the University of Kent in Canterbury, England, criminals may well have a harder time hiding their faces. The software uses interactive genetic algorithms to help witnesses remember details and thus helps police departments in obtaining better sketches of suspects. The software is being used by approximately 15 police departments in the U.K. and by about half a dozen European countries including France and Switzerland.