Personal Authentication Using Finger Images

Abstract

As the need for personal authentication increases, biometrics systems have become the ideal answer to the security needs. This paper presents a novel personal authentication system which uses simultaneously acquired finger-vein and finger texture images of the same person. A virtual fingerprint is generated combining these two images. The result of the combination i.e. the virtual fingerprint is then subjected to pre-processing steps including binarization, normalization, enhancement and Region of Interest (ROI) segmentation. Gabor filter is used to extract features. The feature extracted image is matched with the database. This proposed system is designed such that to achieve better performance in terms of matching accuracy, execution time, memory required and security.
DOI: 10.17762/ijritcc2321-8169.150177