Cancelable Biometrics A Case Study In Fingerprints Of The Gods
- Anil K. JainMichigan State UniversityVerified email at cse.msu.edu
- Rama ChellappaProfessor, ECE, Affiliate Professor, CS, UMD, College ParkVerified email at cfar.umd.edu
- Venu GovindarajuSUNY Distinguished Professor, State University of New York, BuffaloVerified email at buffalo.edu
- Vishal M. PatelA. Walter Tyson Assistant Professor, ECE, Rutgers UniversityVerified email at rutgers.edu
- Gaurav AggarwalResearch Assistant Professor, University of Notre DameVerified email at nd.edu
- Jinyu ZuoPrincipal Research Engineer, SymantecVerified email at mix.wvu.edu
- Jaishanker PillaiSoftware Engineer and Researcher, Google Inc.Verified email at umiacs.umd.edu
- Vivek TyagiResearch Scientist, IBM Research India, IEEE Senior MemberVerified email at eurecom.fr
- Rogerio FerisResearch Manager, IBM T. J. Watson Research CenterVerified email at us.ibm.com
Our fingerprints clearly stand out in the list of different identifiers as they are the most easily accessible. Those curly patterns we see on all of our fingers are an evidence of our uniqueness in this world having a population of around seven billion humans. Every person has a different ridge and valley pattern on their fingers. Observe your hand and try to analyse the special fingerprint pattern. Ridge lines may be entering your finger from one side and taking an exit from the same side after a U-turn, known as a Loop pattern. They may be forming a circular shape at the centre of your finger, known as a Whorl. And the third pattern is known as a Curve in which the ridges enter from one side leave your finger from the other side after a slight bend.
Other than these fairly visible fingerprint characteristics, there are some minutiae features – bifurcation, short ridge, and ridge ending – which can make a big difference between two fingerprints. Analysing these features can be very helpful for the fingerprint scanner to distinguish patterns of people who belong to the same family. Scientific research has concluded that members of one biological family have similar fingerprints, credits given to DNA.
How Does A Fingerprint Scanner Work — The Application Of Biometrics
The fingerprint recognition involves snapping an image using a fingerprint scanner which is then digitally compared with a previously stored snapshot of your fingerprints. The scanner processor sees whether the loops, whorls and curves have any similarity. An important point to be noted is that the scanner doesn’t keep the complete image of your fingerprint. It only stores information about specific points in the form of binary numbers.
Many fingerprint recognition technologies have been developed but only a few of them have seen a widespread implementation. Let’s have a look at these:
Optical Fingerprint Scanners:
These type of scanners use visible light to take a photo of your fingerprints. In this, LEDs are used to illuminate a finger kept on a glass plate. The light reflected from the finger falls onto a Charged-Coupled Device (CCD) present in the scanner. A CCD – also used in camcorders and digital cameras – is basically an array of pixels which respond to falling light over them and generate proportional electrical signals. These signals are then processed to create a digital imprint of your finger known as a “live scan”.
The inverted image so created by the finger sensor represents the ridges – elevated regions – as dark coloured and valleys – depressed regions – as light coloured. You can think of it as a black and white image. One major issue that hampers the quality of the scan is that if the fingerprints are erased, lost due to burns or if the fingers are dirty. This would degrade the accuracy of the fingerprint scanner. Other types of sensors like Capacitive and Ultrasonic provide a solution for this problem. The final image generated should have appropriate darkness level and adequate sharpness to get qualified. If it fails to meet the standards, the image is discarded and sensors settings are adjusted to get the appropriate image in the next shot.
Capacitive Fingerprint Scanners:
Taking advantage of the fact that your fingers could be electrified, these type of fingerprint sensors induce an electric charge on your fingers. Don’t worry, you won’t get shocks. Semiconductors chips are fitted inside these scanners. These chips have tiny cells inside them which are made up of capacitors and an inverting amplifier as a part of an electrical circuit known as the integrator.
Take a look at the figure, the triangle shaped thing is the inverted amplifier which has a non-inverting terminal connected to the ground and an inverting terminal connected to a reference voltage. The inverting terminal is also connected to a feedback loop which has two capacitor plates connected (you can see at the top of the circuit). These capacitor plates store charge in them which is accompanied by your finger acting as the third capacitor.
The human skin is gifted with a cell structure that acts an insulating layer between the ridges and the capacitor. For the valleys, the air filled in the gap behaves as an insulator. Insulator present between the capacitor plates is a non-conductive material i.e. it restricts the flow of electric current through it.
Each cell in the semiconductor is approximately the size of an individual ridge’s thickness. The sensor first turns voltage of the system to zero by switching off the reset switch for each cell. This is done to “even out” the voltage across the circuit. Then, the sensor applies a specified voltage value to the circuit after opening the reset switch for each cell.
The voltage applied by the sensor charges the capacitor plates. The third capacitor i.e. our fingers could be moved above the sensor, thereby, altering the total value of the capacitance of the conductor plates. The value of capacitance between the ridges and the conductor plate will more be than that between the valley and the plate, why? This is because if the distance between the capacitor plates is decreased the capacitance increases. We can’t join them together, by the way.
There is a sure shot difference between the capacitance below the ridges and the valleys which help the scanner to detect whether the region above a cell is a ridge or a valley. Following similar steps – analysing capacitance for the complete array of cells – the fingerprint scanner figures out a general pattern of our ridges and valleys helping it to come up with an image of our fingerprint.
During the authentication process, the newly obtained fingerprint is compared with a previously stored version in order to provide access to a device or a building. One advantage of the Capacitance finger scanners over the Optical ones is that it can’t be fooled using a visual impression of your fingerprints, a real finger is required. The worst case scenario would be someone chopping off your finger.
Ultrasonic Fingerprint Scanners:
These scanners work on similar technology used by the ultrasonologists to get a sneak-peak into a mom-to-be’s baby bump. The piezoelectric transducer fitted inside the scanner throws ultrasonic pulses – sound waves having the frequency higher than the human audible range – on the finger placed above it. The waves reflect back to the sensor which are then analysed to create a digital image of the fingerprint.
I told you earlier, your god gifted skin is designed to hold a charge. But a worthy addition to this is that your fingerprint characteristics are present deep inside your fingers – on the dermal layer of your skin – which helps the Ultrasonic scanner to get your finger photos even if your prints are rubbed out, scratched or dirty. This is contrary to what we’ve seen in the case of Optical scanners.
It is obvious that the reflected ultrasonic wave would have different intensity level than the one transmitted, mostly reduced as some amount of the pulse gets absorbed into the skin. The intensity levels are different for waves reflected from ridges and valleys. The Ultrasonic sensor calculates the change in intensity level for reflected pulses from different points on the finger and creates an image of your fingerprints.
The Le Max Pro smartphone houses an ultrasonic sensor known as Sense ID. Manufactured by the Snapdragon-maker Qualcomm, the finger sensor stores all of its data on a separate chip included in the device which acts as a preventive measure for stopping evil minds from stealing confidential data.
Other Aspects of fingerprint scanners:
Fingerprint recognition is the most prominent application of Biometrics. It has been since the year 1891 when an Argentine anthropologist Juan Vucetich used to collect fingerprints of criminals. The pre-historic origins date back to 200 BC, when China’s Qin Dynasty used to keep handprint records as an evidence during the burglary investigation. Juan’s fingerprint archive helped in successful identification of Francisca Rojas as the killer in an 1892 double murder case. Rojas had killed her two children before cutting her own throat to put blame on some outsider.
During the early stages of fingerprint recognition, manual comparison of fingerprints was done for identification of the prints. The biggest advantage of having fingers (and fingerprints) is that you can’t forget them and no one can guess them either, unlike the complex passwords you set for your email accounts. Hopefully, your prints don’t get rubbed off or peeled by some psychopath. That would be the worst way to lose your fingerprints and maybe your fingers also.
Optical scanners are the easiest to fool, just by showing an image of your fingerprints. Not literally, but in the Discovery Channel show MythBusters, the hosts Jamie Hyneman and Adam Savage managed to get a sensor acceptable photographic form of the fingerprints taken from a hand. Capacitance scanners do offer some added advantage but one could use a mould of your fingers to get past the security system. Modern finger sensors do have heart rate and pulse detectors which enhance the level of security. Anyways, if a person manages to steal your unique ridge-valley pattern somehow then all you can do is hope for the best.
A concept of “Cancelable biometrics” was proposed by N. K. Ratha, J. H. Connell, and R. M. Bolle in their research paper “Enhancing security and privacy in biometrics-based security systems” published in IBM Sytems Journal in 2001.
So, a fingerprint recognition system allows only the authorised users to gain access to a device or premises rather an authorised ID card which can be misused by anyone. RFID Tags can also be used to address needs of such kind, but for that, you would have to get a glass-capsuled RFID chip injected in your hand.
The ease of access facilitated by Biometrics is commendable. In the case of a fingerprint scanner, all we have to do is swipe our finger on the sensor. But the question about privacy and safety of the users is a matter of concern. We can’t turn deaf ears towards a matter of sensitive nature. A person who takes advantage of Biometric authentication devices might have to pay some irreparable damages which can’t be reversed using the money.
You’ll find these interesting: