Low Cost Multimodal Biometric identification System Based on Hand Geometry, Palm and Finger Print Texture

Abstract

This paper presents a multimodal biometric identification system based on the combination of geometrical, palm and finger print features of the human hand. The right hand images are acquired by a commercial scanner with a 150 dpi resolution. The geometrical features are obtained from the binarized images and consist on 15 measures. A support vector machines is used as verifier. The palm print and finger texture are obtained by means of different 20 Gabor phase encoding schemes. A robust coordinate system is defined to assure the image alignment. A Hamming distance and threshold are used for verifying the identity. A feature, score and decision level fusion results have shown the improvement of the combined scheme.