Fingerprint recognition, gender classification based on minutiae points in biometrics

Abstract

Fingerprinting remains the best to establish personal identification and tracking criminals. Few researchers addressed the use of fingerprint for gender identification which will be more helpful in short listing the suspects. In this paper, a novel method is proposed to estimate gender by analyzing fingerprints using fast Fourier transform (FFT). A dataset of 400 persons of different age and gender is collected as internal database. Initially all the fingers of the subject were tested . Frequency domain calculations are compared with predetermined threshold and gender is determined. Of the samples tested, 94.5 percent (189 samples are identified exactly out of 200 male samples) and 96 percent (192 samples are identified exactly out of 200 female samples).