The PLR-DTW method for ECG based biometric identification
There has been a surge of research on electrocardiogram (ECG) signal based biometric for person identification. Though most of the existing studies claimed that ECG signal is unique to an individual and can be a viable biometric, one of the main difficulties for real-world applications of ECG biomet...
Saved in:
Published in | 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society Vol. 2011; pp. 5248 - 5251 |
---|---|
Main Authors | , , , |
Format | Conference Proceeding Journal Article |
Language | English |
Published |
United States
IEEE
01.01.2011
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | There has been a surge of research on electrocardiogram (ECG) signal based biometric for person identification. Though most of the existing studies claimed that ECG signal is unique to an individual and can be a viable biometric, one of the main difficulties for real-world applications of ECG biometric is the accuracy performance. To address this problem, this study proposes a PLR-DTW method for ECG biometric, where the Piecewise Linear Representation (PLR) is used to keep important information of an ECG signal segment while reduce the data dimension at the same time if necessary, and the Dynamic Time Warping (DTW) is used for similarity measures between two signal segments. The performance evaluation was carried out on three ECG databases, and the existing method using wavelet coefficients, which was proved to have good accuracy performance, was selected for comparison. The analysis results show that the PLR-DTW method achieves an accuracy rate of 100% for identification, while the one using wavelet coefficients achieved only around 93%. |
---|---|
ISBN: | 9781424441211 1424441218 |
ISSN: | 1094-687X 1557-170X 1558-4615 |
DOI: | 10.1109/IEMBS.2011.6091298 |