BIOMETRICS IMPLEMENTATION, COMPARISON, FINGERPRINT SCANNERS
In recent years, biometric identification systems and fingerprint recognition systems, in particular, have been widely adopted by both government as well as private outfits. Governments across the nations have been using this technology for the purposes like civil identity, law enforcement, border control, access control, employee identification, attendance, etc. Business setups have been using it to save time by streamlining various processes like employee identification, physical and logical access control, user authentication, safeguarding cloud communication, etc. Biometric systems have been embraced by organizations of all sizes and shapes regardless of their industry type and vertical. Availability of fingerprint sensors in affordable mobile devices and government national ID programs have particularly brought biometrics to a common man and have increased awareness as well as acceptance. Biometric systems are also getting more and more inexpensive due to widespread implementation and increasing rate of adoption.
Evolution of biometric systems over generations
Early generations of biometric devices were not as efficient as modern ones. They were bulky, heavy and required supervision during operation. They were also not as fast as current devices and required calibration for accuracy. PC integration was not available in the first generation biometric devices and usage were limited to law enforcement applications. The second generation of biometric devices brought some improvements over the first generation but was still expensive and had high FRR and/or FAR. Finger preparation was also required prior to scan as sensors were not as technologically advance as modern sensors. Only optical sensors were available in recognition systems, other sensing methods were either unavailable or under development. For second-generation biometric systems, applications were limited to high-security computing in vertical applications and building access control.
Two different types of modern fingerprint scanners. Left: USB Fingerprint scanner requires a PC or smartphone with supporting the application to work. Right: Commonly used for access control, can scan and store many templates and transactions independently, can connect to a PC or network to communicate and export biometric data. The current generation of biometric systems is available with sensors leveraging different techniques like capacitance, thermal, etc. to read fingerprints. They come with the ability to detect liveness and do not require manual calibration.
They come with the ability to detect liveness and do not require manual calibration. Current biometric recognition systems are considerably faster than earlier generations. They have SDKs available for PCs and come with encryption support. Mass production induced by increasing adoption rate has not only slashed prices of biometric systems but also encouraged their usage in the mainstream identification and authentication methods. Now billions of people use biometric identification and authentication in some or other way on a daily basis. From unlocking doors or unlocking phones, biometrics is always at work.
Increasing adoption rate – decreasing prices
Mass production cuts down prices, and that is what exactly happening with biometric recognition systems right now. Increasing numbers of implementation made mass production of biometric systems imperative and slashed prices. A biometric system price may depend on factors like brand, certifications, waterproofing, type of sensor, etc. A small USB fingerprint scanner can cost as little as $50 and a sophisticated ten finger scanner with live finger detection ability can cost $2500 as well. Increasing production and completion are expected to lower the prices further. An average selling price of global mobile fingerprint sensor volumes is estimated to be dropped to $2 per unit in 2020 which was as high as $5.5 in 2014.
Increasing adoption has helped bring down the cost of biometric devices. Factors like an economy of scale, increasing production and electronic components getting cheaper, have helped biometric systems to become affordable for small business and even for individual applications. Slashing prices are particularly evident in the case of fingerprint scanners. Earlier, only high end or flagship mobile phones were equipped with fingerprint sensors, but now even a $100 smartphone offers a capacitive fingerprint sensor. Fingerprint recognition systems, which were earlier used only in high-security facilities or restricted areas, are now commonly seen everywhere. Let it be office doors, server rooms, schools, banks, POS, etc., fingerprint scanners have made their way to everyday life. Due to mass production, building blocks of biometric systems are getting cheaper and new entrants are offering very competitive prices. Technological enhancements and the introduction of new hardware also slash prices of previous iterations.
There are various elements to consider before choosing any modality to employ for a biometric application. Level of security required, cost of the biometric system, return on investment, etc., are some of the elements that may become deciding factor in employing a biometric recognition method.
Following table compares different aspects like accuracy, cost, template size, stability and level of security of common biometric recognition methods:
Fingerprinting: the biometric method of choice
Different biometric recognition methods offer the different set of features, advantages, and disadvantages. Cost is also an important factor to consider while choosing a biometric recognition system. For high-security applications, multi-factor authentication or multi-modal biometric implementation can be considered, while low-security applications can be implemented with a single biometric modality. Multi-modal biometric applications may hike up the investment required multi-fold, so there has to be a balance of everything and a thorough return on investment study may be required before taking up multimodal biometric recognition. Fingerprinting is the most popular modality among all biometric recognition methods. Being inexpensive, easy to implement and use, it has most penetration in authentication and access control applications as well as consumer electronics like mobile phones and portable devices. Fingerprint scanners make use of sensors to scan a pattern. These sensors come equipped with different techniques to read and produce the image of the fingerprint pattern.
Optical sensors capture the image of fingerprints with a specialized digital camera setup. This is the most common type of fingerprint sensors, which are widely available at cheap prices. Optical sensors pose shortcomings like the quality of the scan is impacted by dirty fingers and they are easier to be tricked than other types of sensors.
Capacitive scanners make use of pixel array of capacitors instead of visible light, to produce the image of fingerprints. Capacitive scanners are hard to forge because they cannot be fooled by fingerprint images. They are more expensive than optical sensors.
Ultrasonic scanners use the very high-frequency sound wave to read the pattern of fingerprints. Ultrasonic sound waves reflected from the fingertip surface are measured by the sensor and fingerprint pattern image is produced. Performance of ultrasonic sensors stays unaffected by the dirtiness of finger surface as it doesn’t capture the image like optical sensors.
Thermal line sensors
These sensors read a fingerprint pattern by measuring temperature variation in fingertip ridges and valleys. It requires the finger to be moved over a linearly arranged narrow array of thermal sensors. They are small in size and require finger movement to measure fingerprint patterns.
Cost of a fingerprint recognition system can be highly dependent on the sensor type used in the device.
Use of fingerprint scanners rose steeply during the 2010s. Consumer electronics, especially mobile phones and tablets made extensive use of fingerprint sensors.
Following chart shows how revenue and order inflow skyrocketed for fingerprint cards from Q3 (2014) to Q4 (2015):
Both businesses and governments have recognized the potential of biometric recognition systems and are leveraging them for various identification and authentication purposes. With successful adoption at various fronts like access control, civil identification, border control, law enforcement, etc. it can be said that biometrics is rapidly growing and has good prospects for the future. Global adoption and successful implementation across industries have shown that biometrics is the way forward. Market intelligence companies also predict exponential growth of biometric recognition in the future.
Newer trends like cloud biometrics is set to take the biometric affordability to the next level. According to an estimate by Frost and Sullivan, market revenue for fingerprint authentication on mobile devices is to increase from US$52.6 million in 2013 to US$396 million in 2019. Authentication with BaaS on mobile devices can safeguard crucial operations like banking transaction, authorization of payments, e-commerce transaction, etc. In banks, scalability is an important aspect to take care of. Launch of new branches, installation of ATMs, etc. are a part of ongoing banking operations. Biometrics as a cloud service can benefit scalability intensive industries. New ATM with biometric capabilities will not only be securer than traditional card and PIN based authentication, but also be easy to deploy with cloud biometrics.
Biometrics have proved to be more efficient, faster and secure than traditional identification practices like ID cards, access cards, PINs and passwords, which are either possession based or knowledge-based authentication factors. Biometrics, being an inheritance based factor, eliminates the possibility to forget or share passwords and loss or theft of ID/access cards. Implementing multimodal biometrics or multi-factor authentication with biometrics as one of the essential factors provides even greater security, which is a common requirement in high-security facilities like military setups, data centers, nuclear reactors, R&D facilities, etc. This ability makes biometrics to be offered as a service over the internet. Huge success and adoption of biometric devices have induced mass production and these devices.
Growing adoption of biometric recognition systems across all industries and sectors in a variety of applications has paved the way to a huge commercial market for devices and solutions. Modern biometric systems have the ability to connect to information systems and the network. They can share data with a remote server with the centralized biometric database over the internet. Once expected only in high-security facilities, biometric recognition systems have reached in the pocket of the common population. Today, biometric recognition systems have come to the price point, where small businesses and even individuals can easily afford them for office/home security, attendance, employee/customer identification, membership management, the point of sale, etc.