Ajitha T Abraham and Asst. Prof. Yasim Khan M
College Of Engineering, Kottayam, India

This research implements a novel and simple method of age classification using fingerprints. Two methods are combined for gender classification. The first method is the Singular Value Decomposition (SVD), employed to extract fingerprint characteristics by doing synthesis and reconstruction. The second method is the analysis for feature extraction by using 2D Bi-orthogonal Wavelet decomposition, up to 4 level decomposition used for the process of gender identification. This method is experimented with the internal database of 250 fingerprints finger prints in which 125 were male fingerprints and 125 were female fingerprints. Tested fingerprint is grouped into any one of the following five groups: 6-7, 8-12, 13-15, 16-19, 20-30, 30-50 and above 50. Overall classification rate of 60% has been achieved. Results of this analysis make this method a prime candidate to utilize in forensic anthropology for age classification in order to minimize the suspects search list by getting a likelihood value for the criminal gender.

Fingerprint, SVD, Wavelets,BWT