Overview of the Last Article
Our last article reviewed the Biometric KPIs and standards which have been set forth and established by both industry and government. Although this area of Biometrics does not get the limelight like the hardware aspect does, it is still a very important aspect.
For instance, if a business or a corporation wishes to procure and implement a Biometric system, there are many factors which have to be taken into consideration. Apart from cost, the management team first and foremost has to decide which particular modality will best meet their security needs.
This is where the role of the KPIs will come into play. Five key metrics were examined. Whenever a Biometrics vendor manufactures a new technology, they often use these same KPIs to benchmark them against other modalities.
But, a potential customer must evaluate these vendors based KPIs against a set baseline, so that they can truly ascertain the effectiveness of the modality they are considering procuring.
Regarding standards, three major groupings were also examined, which are as follows:
- The Biometric Data Interchange Formats;
- The Common Biometric Exchange Format Framework;
- The Biometric Technical Interface Standards.
In general terms, these standards set the guidelines and best practices which create Biometric Templates into a standard format, which thus can be communicated to and understood by the differing modalities.
Also, they establish the “universal language” for the metadata which is associated with the Enrollment and Verification Templates. Finally, these standards also set forth the degree of communications which is required between the Biometric device, the software applications, and the middleware.
This article examines into much greater detail the often forgotten about a component of Biometric Technology-the sensor. It is considered to be at the “heart” of any modality. This is so because the sensor can be considered as the beginning of the pipeline in the Verification and/or Identification of an individual.
After all, it is from here where the raw images of our physiological and behavioral traits are captured, and the unique features extracted from.
An Overview into The Biometric Sensors
What exactly is a Biometric Sensor? It comes in many different shapes and sizes. The type of Sensor technology used with the various Biometric modalities can be quite different from one another. For example, with Fingerprint Recognition, an optical based Sensor is used, and Facial Recognition systems make use of special types of cameras. With Keystroke Recognition, the computer keyboard itself is the Sensor, and with Voice Recognition, specialized microphones or even a Smartphone can become the actual Sensor.
A Biometric Sensor can be specifically defined as “. . . hardware . . . that converts Biometric based input into a digital signal and conveys this information to the processing device” (SOURCE: 1). In the world of Biometrics, there are two types of Sensors:
Within these two groupings, they are many types of scanning technologies:
Optical based scanners;
Solid State Sensors;
Temperature differential Sensors;
Multispectral Imaging Sensors;
Touchless Fingerprint Sensors;
Other types of Sensors (which include CCD Cameras, Active Sensors, and Passive Sensors).
Optical based Scanners
This is the most commonly used Sensor with live scan Fingerprint Recognition Technology. Live scan simply means that that a finger with a pulse is required for Verification and/or Identification. For those applications which make use of Fingerprint Recognition, a platen is normally used.
It is made of a glass-based composite, and from underneath it, a Light Emitting Diode (LED) is beamed at an upward angle. This is then reflected back onto a Charged Coupled Device, also known as a CCD. A CCD is merely a camera which can transpose light into electrons.
Thus, as the finger is placed on the platen, the light captures the ridges of the fingerprint back into the CCD. At this point, the ridges appear as dark lines, and the valleys and whorls simply appear as white spaces. Optical based Sensing technology possesses a number key advantages such as:
- Low cost
- A relatively high resolution
- Strong ability to take into account various temperature changes.
This type of Sensor technology is actually proving to be a very good alternative to the traditional Optical Scanner and is the second most widely used Sensor in Fingerprint Recognition. However, rather than using a CCD, Solid State Sensors utilize an array of electrodes to capture the image of the ridges of the fingerprint.
A capacitance level is formed between the ridges of the fingerprint and the electrode. From this, the raw image of the fingerprint is then created from which the unique features can be extracted from.
There is also another type of Solid State Sensor which is known as a “Sweep Sensor.” This contains a very few number of electrodes, so the fingertip of the individual must be literally swept from left to right (or even vice versa) to capture the full image of the fingerprint.
Solid State Sensing Technology has three primary advantages:
- A much lower cost;
- Very low power consumption;
- A very miniature size (thus it will work very well for use with Smartphones and other types of wireless devices).
This type of technology works in very much the same way as the ultrasound machines used in the hospitals and doctor’s offices. From the Ultrasound Sensor, acoustic pulses are sent to the fingerprint, and a receiver from within the Biometric device then captures the returning acoustic pulses. This results in a raw image of the fingerprint being created.
A key advantage of Ultrasound technology is that is that the actual creation of the raw image does not depend on upon any type of visual capture. Thus, it is not prone to degradation like the Optical Sensors.
Temperature Differential Sensors
This form of sensing technology relies upon the pressure differences to create the image of the fingerprint. With this, electric currents are created and then transmitted to the ridges and the valleys of the fingertip. As a result, different electric intensities are created. This is known as specifically as the “Piezoelectric Effect.”
With Temperature Differential Sensors, a pyroelectric material is utilized to generate the varying levels of electric current. Thus, the differences in the temperature of the fingertip can be gauged.
Multispectral Imaging Sensors
As the name implies, Multispectral Imaging technology produces numerous raw images. To do this, varying wavelengths and differing levels of polarizations are utilized. These are then combined into one overall image of the physiological trait.
This particular technology was created to help compensate for a major weakness of the Optical Sensing Technology. That is, if there are any objects embedded into the physiological trait itself (such as a fingertip), this could also be captured. By using a Multispectral Imaging Sensor, this issue is virtually nonexistent.
Ethical Hacking Training – Resources (InfoSec)
Touchless Fingerprint Sensors
Once again, as the name implies, there is no direct contact or interface, which is required by the end user. There are three specific types of touchless technologies:
This shines a light onto the physiological trait from different angles using just one type of camera. In this case (if Fingerprint Recognition is being used), the images from both the ridges and valleys are captured, unlike the full contact Sensors. A primary disadvantage of this type of Sensor technology is that the end user must keep his or her finger absolutely steady. Also, this technology must come into compliance with the best practices and standards as set forth by the FBI.
A special red light is shone through the physiological trait. With Fingerprint Recognition, this light is aimed towards the sides of the finger.
Three Dimensional based:
With this, either parametric modeling or nonparametric models are used. With the former, the raw images of the physiological trait are projected onto a cylindrical model. With the latter, mathematical algorithms are used to model the unique features of the physiological trait. As a result, much more granular detail can be captured.
Other Types of Sensors:
There are also other types of Sensor Technologies which are available, but these are primarily used in Facial Recognition, and the Behavioral based Biometric modalities. This includes:
CCD Cameras (also known as CCDs):
CCDs are cameras and can yield pictures of the face from surveillance video. The output can be black and white and even color, and can also work with other spectrums of light, not just those in the visible light range. With CCD sensors, multiple images of the face are taken in a rapid-fire succession and then are compiled into one facial image from where the unique featured can be extracted.
With this type of Sensor technology, typically only one camera is used to take the various pictures of the face. A special light is then used to project a specific pattern (such as a geometric plane) onto the face. The differences captured from this special pattern allows for a 3-D image of the face to be constructed.
With this, multiple cameras are utilized, and the 3-D images of the face are then constructed through a process known as “triangulation.” Typically, in these scenarios, two calibrated cameras are used to take pictures of the face from the left to the right, and vice versa.
As it has been reviewed in this article, the Biometric Sensor is at the heart of any modality. Being that it is such a crucial component, it is very important that it is constantly monitored for any signs of physical damage, dirt issues, or even technological failures. The sensor must be maintained at an optimal performance level 24 X 7 X 365.
The trend is now to start using non-contactless Sensors with all of the Biometric modalities. This is not only convenient from the standpoint of deployment for the business or corporation, but also, it will help to greatly increase the user acceptance of it, primarily because of hygiene-related issues.
The technology behind Biometric Sensors is continuing to grow at a very rapid pace. There is strong movement within the Biometrics Industry to make them as miniature as possible so that they can be used on wireless devices and Smartphones. In fact, Apple has already adopted the use of Fingerprint Recognition for their iPhone 5 and later models.
At the very bottom of the device, one can notice a circular object. This is, in fact, the Fingerprint Recognition Sensor. To deploy this technology, Apple bought out a leading Biometric Sensor manufacturer known as “Authentic.” Apple has patented this technology and is planning to use it further in their other wireless devices as well.
- Certified Biometrics Learning System, Module 1, Biometrics Fundamentals, © 2010 IEEE, p. 1–28.