Title :
ECG based biometric for doubly secure authentication
Author :
Safie, Sairul I. ; Soraghan, John J. ; Petropoulakis, Lykourgos
Author_Institution :
Centre for Excellence in Signal & Image Process. (CeSIP), Univ. of Strathclyde, Glasgow, UK
fDate :
Aug. 29 2011-Sept. 2 2011
Abstract :
This paper presents a new ECG based biometric system that has a doubly secure characteristic when applied to authentication. In this system the parameters of the new underlying pulse active ratio (PAR) feature vector process are represented by a 4 digit security personal identification number (PIN). An authentication is achieved by firstly using the correct PIN and, secondly, by obtaining a positive ECG feature vector matches. The proposed method is validated by experiments on 113 subjects recorded over wide time duration. It is shown that the PAR parameters can be used to generate different set of features with similar high authentication performance in a doubly secure fashion.
Keywords :
biometrics (access control); electrocardiography; vectors; ECG based biometric system; PAR feature vector process; PIN; doubly secure authentication; personal identification number; positive ECG feature vector match; pulse active ratio feature vector process; Authentication; Biometrics (access control); Databases; Electrocardiography; Feature extraction; Vectors;
Conference_Titel :
Signal Processing Conference, 2011 19th European
Conference_Location :
Barcelona