Feasibility of a wearable band type biometric devices using skin spectroscopy

Abstract

In this work, we studied the feasibility of MSP (Multispectral Skin Photomatrix) module as a new wearable biometric devices. The MSP device measures optical patterns of the wrist skin tissue. Optical patterns consist of 2×16 photocurrent intensities of photodiode arrays, which is generated by optical transmission and diffuse reflection of photons from LED light sources with variable wavelengths into the wrist skin tissue. Optical patterns detected by the MSP device provide information on both surface and subsurface characteristics of the human skin tissue. Totally 73 subjects have been participated in this study, they showed their unique characteristics, as determined by several wavelengths of light. The experimental results show that the best personal identification accuracy can be acquired using a combination of infrared light and yellow light. This novel biometric device, the MSP module, exhibited an excellent false acceptance rate (FAR) of 0.292% and a false rejection rate (FRR) of 3.539%, which are similar value of another biometric technologies such as voice or face recognition, etc. From these experimental results, we found that people exhibit unique optical patterns of their inner-wrist skin tissue and this uniqueness could be used for developing novel highaccuracy personal identification devices.