A Comparative Study of Gender Classification using Fingerprints
Shadab Alam, DIPTI DIPTI, M.R. Dua, Ashutosh Gupta · International Conference on Computing for Sustainable Global Development · 2019
Fingerprints are one of the primary biometric traits that are the most reliable and the legitimate source of evidence in the court of law. They can be used to determine the gender of the person based on the latent fingerprint obtained from the crime scene. The research work is influenced by studies of Psychology and Anthropometry which suggests that gender can be classified based on the structural analysis of the geometric features present in the hands. Global characteristics such as white lines count, ridge thickness valley ration (RTVTR) and ridge density are obtained using the ridge density as a feature in gender determination. It uses the support vector machine as the classifier to determine the gender of the person. This paper proposes a novel method based on singular value decomposition (SVD) and discrete wavelet transform (DWT). The database consists of 420 fingerprint samples collected from males and females of age group ranging from 14 to 60 using the capacitive sensor. The algorithm is tested on the collected database for gender determination.