Evaluation on palm-print ROI selection techniques for smart phone based touch-less biometric system

Abstract

There are many methods have been carried out for human recognition such aspersonal identification number (PIN), password or ID card but all of these methods can beguessed, hacked or stolen. Palm-print verification system is a biometric technology which isdeveloped to authenticate person based on individual palm-print pattern. This paper presentsan initial effort to perform touch-less palm-print recognition system by considering theeffective way to extract the palm-print region of interest (ROI). The system starts with handimage collection using smart phone device. This project proposes two hand trackingalgorithms i.e. two point method and canny method so as to detect the peak and valley of thepalm. Afterwards, the desired ROI is selected and the palm-print ROI is stored in database forthe evaluation of their appropriateness to be used for the touch-less palm-print recognitiondata.