miBEAT Based Continuous and Robust Biometric Identification System for On-the-Go Applications
In recent years, Biometric identification has taken a giant leap from objective security access system such as retina scan or a finger print scan to a continuous biometric identification based system and for that a single lead Electrocardiogram (ECG) sign
- PDF / 338,185 Bytes
- 7 Pages / 439.37 x 666.142 pts Page_size
- 101 Downloads / 186 Views
Abstract In recent years, Biometric identification has taken a giant leap from objective security access system such as retina scan or a finger print scan to a continuous biometric identification based system and for that a single lead Electrocardiogram (ECG) signal is considered to be a good marker. However the parameters normally considered for biometrics from ECG normally requires several parameters which again depend on a good resting signal. For applications involving on-the-go Biometric identification, such systems do not provide a reliable solution. This paper describes a novel approach to a robust and continuous biometric identification system by obtaining touch based ECG as well as Photoplethesmogram (PPG) signal simultaneously from miBEAT (an open source CE certified innovative platform to develop medical grade systems) and by mapping variability features in real time common to both the signals. By validating the system on 20 healthy individuals, it was found that this system works with minimum limitations and thereby can be considered for a robust biometric identification system where higher security measures are required. Keywords Continuous biometric identification Variability
miBEAT
ECG
PPG
J. Yathav (&) A. Bailur Abhinav Cardea Labs, Cardea Biomedical Technologies (P) Ltd, New Delhi 110067, India e-mail: [email protected] A. Bailur e-mail: [email protected] Abhinav e-mail: [email protected] A.K. Goyal Center of Excellence in Biomedical Instrumentation and Signal Processing, Noida Institute of Engineering and Technology, Greater Noida 210306, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2017 N. Modi et al. (eds.), Proceedings of International Conference on Communication and Networks, Advances in Intelligent Systems and Computing 508, DOI 10.1007/978-981-10-2750-5_28
269
270
J. Yathav et al.
1 Introduction According to a report by Global Industry Analysts (Global Industry Analytics, 2011) Biometrics market is estimated to cross $16 billion by 2017. Several distinctive features such as facial features, hand geometry, speech, walking manner, hand writing, prints in finger and iris have been explored to identify an individual in the conventional biometrics systems. Phua et al. [1, 2] validated exponentially uniqueness of the heart, where two heart sounds form different characteristics. Recently scientists, researchers and developers have migrated from such sporadic analytics to a continuous authentication system which involves biomedical signals such as ECG among others. Validation of ECG as the parameter is accompanied by the fact that the geometrical and physiological differences of the heart’s function in different individual indicate uniqueness in the cardiac signal taken [3]. ECG features have been effectively incorporated for high precision identification and authorization of individuals to access the secured portion [4]. However, further analysis using ECG for an individual revealed obvious characteristic that may not be present in rec
Data Loading...
