Research Of Finger-Based ECG Identification Algorithm In Time Frequency Domain

In this highly developed digitalization and information technology era,how to realize identity of individuals safely,rapidly and accurately has become a present question by the public.Biometrics is a growing research in realizing automatic identification using human intrinsic characteristics,such as physiological or behavioral.And it is featured with high accuracy and reliability identification.However,there are still some security issues occurring in biometrics,like fake fingerprint,false iris and so on.As the potential security issues hidden in biometric applications have caught the extra attention of researchers at home and abroad,several new biometric technologies with high security have been proposed intermittently.ECG is the endogenous signal that contains relevant information for human identification,the outstanding performance of which is the realization of live detect identification with high security.Accompanying with the emergence of ECG chips with the features of small size,lower power and easy integration recently,acquisition of the ECG signal from fingers can be realized,which appear to open a great promising prospect for ECG-based biometric identification technology.This paper mainly studies the finger-based ECG identification algorithm in time frequency domain.First,after analyzing the characteristics of the finger ECG,a new finger ECG denoising method based on wavelet threshold and genetic algorithm is proposed.Second,the time frequency distribution of finger ECG is studied,and the finger ECG identification algorithm based on generation S transform and Ziv-Merhav(ZM)cross parasing is put forward.Thirdly,the algorithm proposed in this paper is verified by three finger ECG data and results have revealed that this is a promising method.Finally,the proposed finger ECG identification algorithm is implemented in the android smart phone.The main ideas of this paper are mainly as follows:1.The changes of ECG acquisition and the development of ECG identification algorithm are reviewed,then the mechanism and characteristics of the ECG waveform are explained followed by the introduction of the identify evaluation index and finger ECG database,which lays the foundation for further study of identification.2.Finger ECG preprocessing algorithm is proposed,including de-noising,R peak detection,segmentation and normalization.First,after a systematic study of wavelet transform,wavelet threshold de-noising and generation algorithm,a new finger ECG wavelet threshold de-noising algorithm based on generation algorithm is proposed,which can dynamically get the optimal solution to fit the threshold in accordance with the characteristic of finger ECG.Then the algorithm3.is verified by fitting ECG and finger ECG.Finally,the signals are processed via R peak detection,segmentation and normalization after de-noising.4.A new finger ECG identification based on generalized S transform and Ziv-Merhav cross parsing is put forward.First,the basis theory of S transform,generalized S transform,singular value decomposition and Ziv-Merhav cross parsing is studied.Second,singular values are obtained through the generalized S transform and singular value decomposition after pre-processing finger ECG singals.Features are extracted by quantified the singular value and are classified by classifier constructed by ZM cross parsing algorithm.Finally,this algorithm is verified by three database including CYBHi,Surface ECG data and Finger ECG data.And the performance is tested separately under different conditions of the training time,the number of singular value and the number of quantization,etc.4.The finger ECG identification application based on Android mobile is developed.It contains four basic functional modules including bluetooth connection,user registration,user identification and system management.This application can receive and display finger ECG,process and extract finger ECG feature,realize classification and identification,etc.This paper finishes the study of the individual identification based on finger ECG signals,and provides a theoretical basis and technical support for the practical use of ECG identification technology.