FINGERPRINT BIOMETRIC MATCHING ALGORITHMS AND SOFTWARE
Algorithms are an integral part of a high performance fingerprint matching system. The purpose of the algorithm is to derive high quality match scores through the comparison of a stored fingerprint template to a live touch image. As touch sensors have evolved to increasingly smaller geometries based the limited real estate on today’s smart phones, algorithm complexity has increased substantially in the derivation of security strength with smaller and smaller amounts of information. Traditional algorithms have relied on large images where a set of minutiae features can be readily extracted and matched. As touch sensors continue to get smaller, advanced algorithms become critical in order to provide higher fingerprint accuracy rates, at faster matching speeds, in more operating system environments, with more efficient yields and costs.
IDEX has a range of available algorithms for use in different applications such as mobile and cards. The available algorithms include proprietary algorithms for both Eagle off chip sensor solutions and Cardinal silicon sensors, as well algorithms optimized for the lower compute resources available in Card / Embedded applications.
Our algorithms are tightly coupled to the hardware based image delivery system and highly optimized to minimize the requirements for processing capacity and memory (both RAM and flash). Everything is provided in re-locatable library files that are independent of operating system, compiler and development environment. Trusted execution environment integrations have deployed our algorithms within the required memory constraints achieving trusted architectures for our customers.