ECG biometrics explored for high-tech military and low-cost home applications
A new device has been developed for the Pentagon to enable U.S. Special Forces to identify people with heartbeat biometrics with an infrared laser from a distance of 200 meters, according to MIT Technology Review.
The “Jetson” device uses laser vibrometry to detect the subject’s heartbeat from surface movements through typical clothing, though it will not work through winter coats or other thick clothing items, the report says. It also requires the subject to be fairly still for around 30 seconds. Officials believe it can be made to work from much further than 200 meters, though, and using the device with an algorithm developed by a team led by Pentagon Combatting Terrorism Technical Support Office Program Manager Steward Remaly can yield match accuracy above 95 percent of the time, under certain conditions.
State University of New York at Buffalo researcher Wenyao Xu says cardiac biometrics have great potential compared to facial recognition. “Compared with face, cardiac biometrics are more stable and can reach more than 98% accuracy.”
New research also indicates that ECG signals collected with a consumer-grade monitor could potentially be used for authentication. “A Key to Your Heart: Biometric Authentication Based on ECG Signals” by Nikita Samarin of the University of California, Berkeley and Donald Sannella of the University of Edinburgh shows single-reading error rates of only 2.4 percent, and 9.7 percent error rates for ECG readings collected four months apart.
The researchers collected samples from 49 men and women with a $99 device, Naked Security reports, using an “off-the-person,” or non-contact approach. They found ECG signals have potential as a practical biometric, but face several challenges, some familiar and some particular to the modality.
No standardized dataset for ECG biometric research exists, so direct comparisons of results is not possible, but similar research in some cases likewise indicates match rates consistent with use for authentication with a little more development, and in two cases reported error rates above 10 percent. Another concern for ECG biometrics is that they are not entirely stable over long periods of time. The researchers suggest that capturing and synchronizing the new signal used in a successful authentication could overcome this challenge.
“This work also demonstrates a high potential of using consumer-grade ECG monitors for authentication,” Samarin and Sannella conclude. “The introduction of low-cost sensors allows system designers to embed them into existing access control systems. Nevertheless, more research needs to be done on extracting features from ECG signals obtained from consumer-grade monitors, preventing spoofing attacks and guaranteeing that ECG-based biometric systems are socially accepted by the general public.”
ECG (or EKG) biometric from B-Secur are expected to be featured in 2020 car models for driver authentication and monitoring.