Human biometric recognition systems are essential in the automated personal identification and verification applications market. This market spans the commercial and military sectors to include a myriad of identification applications from secure entrance, to time card clocks, to bio-certified delivery systems, to automobiles that recognize their drivers, to money transaction systems, to weapons that recognize their owners, etc. A Fast Pattern Recognition (FPR) algorithm has been developed and tested that demonstrates exceptional multi-modal performance in fused facial recognition, palm-print recognition, and fingerprint recognition even under degraded conditions. This same algorithm has also demonstrated exceptional performance on voice recognition and is easily configured to operate on any input that exhibits a pattern, e.g. iris scan. The FPR algorithm is not based on the typical approaches employing eigenfaces, minutia, wavelets, or other models and transforms, but uses specialized pattern processing to compare input features against enrolled features within a database. It is envisioned that this algorithm technology (currently a total software-implementation on a Laptop computer) coupled with conventional biometric tools will demonstrate superior performance as a general purpose biometric recognition toolset for commercial and-or military systems requiring multi-modal biometric fusion based personal identification.
Keywords: Pattern Recognition, Biometric, Fusion, Facial Recognition, Fingerprint Identification, Voice Pattern Recognition, Iris Recognition
