3D Ear Biometrics: Acquisition and Recognition

Abstract

Ear is one of the recent biometric traits used to recognize the human. The ear can be considered as a robust biometric feature with cosmetics, age, and facial expressions, which is unlikely with the other biometrics features. The ear has a predictable background and moreover, the ear can be used an effective supplementary biometric trait with various other biometrics. Furthermore, the distinctiveness property of the ear is highly motivated for security related operations. In this paper, a method for 3D ear acquisition is presented, which is used in creating one of the largest 2D and 3D ear database (IIT Indore) and the recognition performance on the acquired data using a 3D descriptor. We have used Signature of histograms of orientations (SHOT) and Binary SHOT (B-SHOT) descriptor as a 3D feature descriptor and matched the query and repository images and obtain a rank one authentication rate of 97.13% and 97.12%, respectively. © 2018 IEEE.