Biran Abdullah, Jeremic Aleksandar
Annu Int Conf IEEE Eng Med Biol Soc. 2020 Jul;2020:5506-5509. doi: 10.1109/EMBC44109.2020.9176325.
In the recent years, the Electrocardiogram (ECG) based biometric identification has been a subject of considerable research interest. In this paper, we present non-fiducial method for ECG-identification using the short time Fourier transform (STFT), and Frechet mean distance-based algorithms to find the similarity between the STFTs of different people. In this study, we select randomly the training and test data of the ECG in order to test the stability of the method. We apply our proposed method on 124 ECG records of 62 subjects from the publicly available ECG ID database from physionet website. Our preliminary results indicate that the Frechet mean based ECG identification has 96.45% average identification accuracy and therefore can be potentially useful in various applications.
近年来，基于心电图（ECG）的生物特征识别一直是相当受研究关注的课题。在本文中，我们提出了一种使用短时傅里叶变换（STFT）和基于弗雷歇平均距离的算法进行心电图识别的非基准方法，以找出不同人STFT之间的相似性。在本研究中，我们随机选择心电图的训练和测试数据，以测试该方法的稳定性。我们将所提出的方法应用于来自physionet网站公开可用的心电图ID数据库中62名受试者的124条心电图记录。我们的初步结果表明，基于弗雷歇平均的心电图识别平均识别准确率为96.45%，因此在各种应用中可能具有潜在用途。
