Online and Continuous Electrocardiogram (ECG) Biometric Syste

Online and continuous biometric system revolves around continuously monitoring identity of subjects using biometrics while using current and past observations only. Electrocardiogram (ECG) signal is prone to noise interference and is a slow signal to acquire. First, we developed an online abnormal electrocardiogram heartbeat detection and removal using one-class Gaussian mixture model of two components. This outlier removal method was implemented in a biometric system and was examined on a 1,012 fingertip acquired ECG signals database. The biometric system had an equal error rate (EER) of 5.94% in comparison to 12.30% in a state-of-the-art approach. Due to eliminating noisy heartbeats, the system may suffer from data imbalance problem, and for that we proposed a method to synthesize data. Data synthesis was based on the assumption that ECG heartbeats exhibit a multivariate normal distribution. When small sample size dataset was simulated and examined in a biometric system, EER of 6.71% was achieved in comparison to 9.35% to the same system but without data synthesis.;It was desired to increase biometric system robustness and design a continuous authentication system; hence, a novel feature extraction and a unique continuous authentication strategy were proposed. One-Dimensional Multi-Resolution Local Binary Patterns (1DMRLBP), an online feature extraction for one-dimensional signals was designed, and it was incorporated with sequential sampling to establish a continuous authentication system. This system adaptively updated decision thresholds and sample size during run-time. 1DMRLBP accounts for observations' temporal changes and has a mechanism to extract one feature vector that represents multiple observations. 1DMRLBP also accounts for quantization error, tolerates noise, and extracts local and global signal morphology. When 1DMRLBP was applied on the 1,012 fingertip single session subjects database, an EER of 7.89% was achieved in comparison to 12.30% to a state-of-the-art work. Also, an EER of 10.10% was resulted when 1DMRLBP was applied to the 82 multiple sessions database. Experiments showed that using 1DMRLBP improved EER by 15% when compared to a biometric system that was based on raw time-samples. Lastly, when 1DMRLBP was implemented with sequential sampling to achieve a continuous authentication system, 0.39% false rejection rate and 1.57% false acceptance rate were achieved.