An approach to a fingerprints multi-agent parallel matching system

Fingerprints databases are heterogeneous and large, also fingerprints matching are a high computational task, and thus there is a need for flexible and independent methods for automatic parallelization of fingerprints matching systems especially for real time applications. In this paper, we describe a strategy for parallelizing fingerprints matching systems using a collection of collaborating matching agents that are distributed on fingerprints database to collect information, matching them with the unknown fingerprint. After their termination, they discuss with each other their outcomes and the agents with high matching scores deliver their results to the matching manager to produce the candidates' list. We believe that multi-agent technology offer exciting possibilities for the design of next-generation fingerprints matching systems