Sensor Format and Size

As fingerprinting is an analogue application, the laws of physics dictate that if you reduce sensor size, you sacrifice security or convenience (or both).

The Fundamentals: 
A fingerprint system captures our prints and locates a number of unique features (minutiae points or other characteristic patterns). From the locations of these points or patterns and their interrelations, the system assesses the probability that a finger presented to the sensor actually belongs to the rightful user of the system.

Real-life Challenges: 
Fingerprinting in the lab differs from real life. In real life, a variety of challenges present themselves. These include:

• Feature poor prints

The effectiveness of fingerprinting depends on all human fingers being different. However, nature does not evenly distribute these distinctions among fingers and persons. Some fingerprints contain a limited number of unique features, whether they have existed since birth or because of permanent damage. Sensors need to be large enough to compensate for feature poor prints.

• Non-uniform placements

In real-world behavior, people do not place their fingers in a fully uniform fashion. Skewed, translated or rotated placements reduce the overlap between enrolled prints and prints presented for matching. Therefore sensors need to be large enough to compensate for the variation in placements.

• Varying finger conditions

In real-life usage, fingers become wet, dry, cut, dirty or worn.  All of these conditions represent lost or false unique features in the user print. So sensors need to be large enough to compensate for these variations in finger conditions.

Dealing with the real-world challenges described above requires a fault tolerant system that includes enough buffer capability to operate in a reliable fashion.

Can Algorithms Compensate for Size ?
Various sensor algorithms have been around for many years. Minutiae-based systems have been the dominant solution in commercial systems and quality critical open mass-market applications. Pattern-based algorithms have been primarily applied to small format sensors.

Algorithms have improved gradually over the last decade, but they cannot change the laws of physics. No algorithm can compensate for the fact that a small sensor sensor will capture a smaller portion of a finger. 

 

Security and convenience are by nature opposed forces.  One cannot compromise on sensor area size and maintain a high level of convenience without sacrificing security.