The heart sound characteristic identity recognition method is good in robustness and not easy to copy

Abstract

The invention provides a heart sound characteristic identity recognition method with good robustness and difficult duplication. The method is The method comprises the following steps of bbased on frequency domain characteristics of heart sound signals. C;arrying out preprocessing on the heart sound signals; taking the heart sound signals can be regarded as a short-time stable signals so that ; according to the method, tthe signals are subjected to analyzed subjected to linear prediction analysis conveniently; extracting , linear prediction cepstrum coefficient characteristics of the signals can be extracted, creating a Gaussian mixture model is established for characteristic parameters of all the heart sound signals in a heart sound library to construct a characteristic template library,and wherein then the characteristic template library is used for searching matching signals of the heart sound signals to be detected. Short-term energy information of heart sound data is added to linear prediction cepstrum coefficient characteristics, a very good experimental result is obtained, the method is not prone to counterfeiting, high in recognition success rate, simple and smooth in algorithm, good in feasibility, small in algorithm difficulty and short in program running time, identity recognition can be completed quickly, and the method has great advantages in actual application and popularization.

Description

technical field [0001] The invention relates to a heart sound feature identification method, in particular to a heart sound feature identification method with good robustness and not easy to replicate, and belongs to the technical field of personal identification. Background technique [0002] The problem of personal identification in daily life can be seen everywhere. With the rapid development of electronic technology, automation technology and network technology, people have entered the digital age. In the traditional personal identification technology, personal certificates or passwords are used to identify identities. Because personal certificates are easy to be lost or forged, and passwords are easy to leak or be forgotten by users, it has been unable to meet the needs of the times. Whether it is an identification method based on a personal certificate or an identification method based on a password, it cannot meet the user's requirements in terms of security, and may ...

AI Technical Summary

Problems solved by technology

[0013] One is that in the existing personal identification technology, personal certificates or passwords are used to identify identities. Since personal certificates are easy to be lost or forged, and passwords are easy to leak or be forgotten by users, etc., it has been unable to meet the needs of the times.

Whether it is an identification method based on personal certificates or an identification method based on passwords, it can no longer meet user requirements in terms of security, and may even cause huge losses to users.

Although the existing technology also has the identification method of password plus account number, which has improved the security performance to a certain extent, and can also solve the problem of password forgetting by changing the password, but constantly changing the password will still cause great inconvenience to the user. Need one or more more secure, efficient and convenient personal identification technologies

[0014] The second is that although the biometric identification technology of the existing technology has achieved preliminary results, some of which have been promoted for a long time, there are still some hidden dangers in practical applications.

[0015] The third is that the biometric identification technology based on heart sound data is still in the early stage of research and development and has not been promoted. The heart sound identification technology was first proposed by Coxon in 2006, but he only used band-pass filtering in the pre-processing The device filters the heart sound data, which seriously affects the recognition effect, and the accuracy of the recognition result is low, and it is impossible to realize accurate and efficient identification

[0016] The fourth is that the biometric identification technology scheme of the heart sound data in the prior art is not mature enough to achieve practical application and promotion, and involves the positioning of the two components of the heart sound data. Because it is manual positioning, it is affected by noise and has a great impact on the recognition effect. Large, there are a lot of complex and time-consuming operations such as related feature extraction, the algorithm is very difficult, and the program runs for a long time, which is not easy to implement and the effect is not good

Images