A low power VLSI compatible approach for retina tree biometric matching

Abstract

Retinal image is one of the robust and accurate biometrics which can be used to authenticate an individual. Feature matching is a key step for any biometric system and its implementation on hardware structures is often challenging due to the required object based processing. This paper presents an approach for retina tree biometric matching which has the capability to be implemented on a low power and high speed VLSI hardware. The key idea behind the presented method is to extract the Gaussian profile of the retinal feature dataset. The proposed technique is evaluated on the public VARIA retina image database. © 2014 IEEE.