SIFT based dorsal vein recognition system for cashless treatment through medical insurance

Abstract

The time has come not to carry any physical personal identity. The person’s identity will be known by his/her biometric features for any purpose exclusively when the person is unconscious and not carrying any identity proof. Vein recognition system is leading biometric trait in terms of flexibility and security. The fake recognition of veins is almost impossible as the feature points lie underneath the skin and cannot be read without the knowledge of a person (if not unconscious). Most important thing about it is that vein recognitionsystem works only in living persons. In this paper, a recognition system is proposed that works on dorsal vein to claim cashless treatment in a hospital which is on panel of medical insurance company. For acquisition of vein image, a NIR camera VF620 of 850 nm wavelength is used. The proposed model is applied on 1000 samples of dorsal vein patterns of 250 persons. The overall performance of the system was observed to be 97.95% with EER (Equal Error Rate)0.0435%.