Efficient Identification Using a Prime-Feature-Based Technique

Identification of authorized train drivers through biometrics is a growing area of interest in locomotive radio remote control systems. The existing technique of password authentication is not very reliable and potentially unauthorized personnel may also operate the system on behalf of the operator. Fingerprint identification system, implemented on PC/104 based real-time systems, can accurately identify the operator. Traditionally, the uniqueness of a fingerprint is determined by the overall pattern of ridges and valleys as well as the local ridge anomalies e.g., a ridge bifurcation or a ridge ending, which are called minutiae points. Designing a reliable automatic fingerprint matching algorithm for minimal platform is quite challenging. In real-time systems, efficiency of the matching algorithm is of utmost importance. To achieve this goal, a prime-feature-based indexing algorithm is proposed in this paper. The technique involves identifying the most prominent feature of the fingerprint and searching only for that feature in the database to expedite the search process. The proposed architect provides efficient matching process and indexing feature for identification is unique.