Detection of Criminal Traces during Forced Iris Authentication

Abstract

Abstract

Although iris recognition system is considered as most robust, hard to counterfeit and the most secure system of biometric authentication. However the existing system fails to detect a forced authentication which might be misused by criminals to unlock the user's account. In this paper we examine the conditions in which a real user is forcibly presented in front of iris scanner on gun point to unlock the account. In this case a significant difference is noted in the area of iris visibility with respect to user's normal iris area visibility. An abnormal eye blink is also detected in forced condition. We successfully design and developed an algorithm to detect such conditions to protect the users from criminals when a user is forcibly presented to an iris scanner to unlock their account. A sample size of 65 volunteers are taken to record the iris authentication in both the conditions i.e. normal with consent of user and forced under without user’s consent. The average size of iris is recorded 10.1 mm while it expands on 13.2 mm (average) in fear when iris is being scanned forcibly by criminals. We conclude that a variation of 2 to 3 mm in iris exposure is a clear biomarker to indicate some presence of criminal traces and take proactive measures to prevent losses.

Keywords: Criminals, iris dilation, pupil dilation, area of iris visibility (AIV), fear detection, purkinje image, kidnapped, gun point, blink rate

Cite this Article

Vijay Kumar Sinha, Anuj Kumar Gupta, Manish Mahajan. Detection of Criminal Traces during Forced Iris Authentication. Recent Trends in Sensor Research & Technology. 2018; 5(3): 10–18p.