A novel, algorithm metadata-aware architecture for biometric systems

Biometric systems have gained tremendous ground over the past years formulating an advancing market and a wide range of challenging technological research fields. Among the main drivers of this context are the promising capabilities of biometric systems for identification and identity verification purposes through a multitude of means (such as iris, face, gait, etc) as well as their profiling potential. From the technological perspective, a wide range of sensors and a wealth of algorithms for feature extraction, pattern recognition and multilevel fusion have been developed, with the evaluation of biometric systems, however, still indicating that required reliability levels and large scale deployment still remain challenging. The main objective of this paper is to present a novel process flow and a resulting advanced architecture for biometric systems which follows the key principles of Service Oriented Architectures (SOA) and allows for system adaptability and scalability; two major requirements for large-scale deployment of such systems. Within this context, a set of new and enhanced components required for the realization of this approach is presented.