基于心电信号的生物特征提取和识别  

Feature Extraction and Identification of Biometric Information from ECG

作　　者：冯玉蓉 陈玮[2] 蔡光跃[3] FENG Yurong;CHEN Wei;CAI Guangyue(Department of Electronic Engineering,Shanghai Technical Institute of Electronics ＆ Information,Shanghai 201411）（2.College of Biomedical Engineering and Instrument Science,Zhejiang University,Hangzhou 310027）（3 Department of Communication and Information Engineering,Shanghai Technical Institute of Electronics ＆ Infonnation,Shanghai 201411)

机构地区：[1]上海电子信息职业技术学院电子工程系,上海201411 [2]浙江大学生物医学工程及仪器科学学院,杭州310027 [3]上海电子信息职业技术学院通信与信息工程系,上海201411

出　　处：《计算机与数字工程》2018年第6期1099-1103,共5页Computer & Digital Engineering

摘　　要：指纹、人脸、掌纹、虹膜、视网膜等人体生理特征目前已经被广泛应用于商用生物识别领域,然而,这类特征同属"静态"特征,容易被复制、伪造。近年来,心电、心音等"动态"的人体生理特征已被证明可用于生物识别。论文介绍了一种提取心电、心音特征参数时所用到的算法,经小波去噪,综合了23种心电波形特征及支持向量机分类算法的生物识别方法。经过训练识别和测试,12名测试者的24个样本中有21个识别成功,准确率为87.5%。Human body physiological characteristics such as fingerprint,face,palm,iris and retina have been widely used inthe field of commercial biometrics,However,these kinds of features are static,which are easy to be copied and forged. In recentyears,the physiological characteristics of human body,such as EGC and heart sound,have been proved to be used in biological rec-ognition. This paper reports a biometric identification algorithm based on ECG signals. In this method,raw ECG signals are first fil-tered by wavelet,and then classified by SVM. In this research,24 datasets collected from 12 subjects are used to test the method.Result show that 21 can be successfully identified,the success rate is 87.5%.

关 键 词：心电信号 小波去噪 支持向量机 生物识别 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]