Scientists at The Hong Kong Polytechnic University (PolyU) have developed a new system that relies on 3D technology to improve both the speed and accuracy of fingerprinting — all without the pressing required of traditional contact fingerprinting.

According to the scientists, the new contactless 3D fingerprint identification system improves on 2D-scan fingerprinting. The new method, according to the team, is more compact and less expensive than other 3D systems.

"We are pushing contactless biometric technology into a new realm of speed and accuracy at an affordable cost," said Ajay Kumar of PolyU. "This system could be used for many applications, including identification, crime investigation, immigration control, and security of access."

While commonly used by law enforcement agencies, traditional 2D fingerprinting identification methods require the rolling and pressing of fingers against hard surfaces, resulting in either partial or degraded images thanks to skin deformations, smearing and slipping.

The 3D sensors remove the potential for such fingerprinting failures by eliminating the need for direct contact. In doing so, the images are improved and accuracy subsequently improves. Additionally, contactless fingerprinting is considerably more hygienic than contact fingerprinting.

Considered the most reliable features of a fingerprint, ensuring that each fingerprint is unique, minutiae points are details of fingerprints that include ridge endings and bifurcations. On average, roughly 40 to 45 minutiae points per fingerprint can be collected. The scientists developed a system capable of identifying minutiae height and orientation in 3D. These measurements are added to the details concerning location and orientation in 2D, multiplying the amount of data generally captured by traditional fingerprinting systems.

Using less equipment than other contactless 3D fingerprinting systems, the technology developed at PolyU relies on a single inexpensive digital camera along with a handful of LED light sources managed via computer. This, along with a series of proprietary algorithms, can serve to identify 3D minutiae characteristics, thereby making it possible to match fingerprints with a rate of accuracy around 97% in a matter of seconds.

For more on the technology, see the article Towards Contactless, Low-Cost and Accurate 3D Fingerprint Identification.