Fingerprint verification in multimodal biometrics

Abstract

The need to ensure security has spurred the growth of biometric verification. Biometric systems using a single biometric trait for authentication purposes have some limitations. In order to overcome this we are investigating multiple sensors that capture different biometric traits. Multimodal biometrics has the potential to overcome these limitations by improving system security levels and increasing the accuracy. This paper focuses on fingerprint verification as part of a fusion of biometric modalities such as voice and fingerprint. Fingerprint verification is one of the most reliable biometric techniques for personal identification. We determined that to merge the fingerprint system with voice verijication we need to develop our own algorithms for fingerprint verification. We describe the design and implementation of a fingerprint verification system that operates in two stages: minutia extraction and minutia matching. An improved version of the minutiae extraction algorithm was identified in literature and is implemented for extracting features from fingerprint images. ©2006 IEEE.