Personal identification method based on optimal period waveforms of photoplethysmography (PPG)

Abstract

The invention discloses a personal identification method based on optimal period waveforms of PPG, and is mainly used for solving the problem that the existing personal identification method of PPG signals is lower in identification rate. The personal identification method based on the optimal period waveforms of PPG comprises the following steps: 1) obtaining the PPG signals of identifiers, and denoising wavelets and carrying out normalization processing; 2) carrying out waveform segmentation on the processed PPG signals, thus obtaining multiple single-period waveforms; 3) selecting optimal single-period waveforms and removing special-shaped waves; 4) averaging multiple optimal single-period waveforms to obtain average period waveforms; 5) carrying out Fourier transform to obtain amplitude spectrum samples; 6) using non-negative matrix factorization to reduce the dimensionality of the amplitude spectrum samples, then generating test feature vectors; 7) matching the test feature vectors with an identifier training feature vector library generated in advance by using a support vector machine, and outputting personal identification results. By adopting the personal identification method based on the optimal period waveforms of PPG disclosed by the invention, the personal identification rate can reach more than 99.6%, the effectiveness and reliability are high and the personal identification method can be used for occasions requiring high personal identification rates of telemedicine.

Description

technical field [0001] The invention belongs to the technical field of information processing, and in particular relates to an identity identification method, which can be used in telemedicine where the identification rate of patient identification is high and the like is required. Background technique [0002] Traditional identification methods, such as personal resident ID cards, IC cards, account passwords, and motor vehicle driver's licenses, have defects such as loss, forgetting, theft, and vulnerability to attacks, while biometrics based on face, fingerprint, and voice features The method has mature technology and high recognition rate, but it is easy to be forged. In order to enhance the reliability and security of biometric technology, on the one hand, a variety of biological characteristics are used to improve the security level of identification; on the other hand, a variety of physiological characteristics are used for identification. [0003] Photoplethysmograph...

AI Technical Summary

Problems solved by technology

[0005] Gu, Y.Y., Zhang, Y.T. used the method of fuzzy decision-making to achieve identity recognition in the article "Photoplethysmographic authentication through fuzzy logic" published at the "IEEE EMBS Asian-Pacific Conference on Biomedical Engineering" in 2003. In the same experimental environment, it can reach 94% Recognition rate, if the human body is in a different environment, due to the influence of breathing, motion artifacts, etc., the recognition rate can only reach 82.3%, which cannot meet the requirements of high identification applications

[0006] Turkey's A. In the article "A novel feature ranking algorithm for biometric recognition with PPG signals" published in the journal "Computers in biology and medicine" by Kemal Polatb et al. in 2014, the K-nearest neighbor classifier was used to complete the identification based on PPG signals, with a maximum of only 94.44% The recognition rate is high, and the feature dimension of the input classifier is large, and the algorithm complexity is high

Images