Biometric apparatus and method using bio signals

• US 7,617,058 B2
• Filed: 11/08/2005
• Issued: 11/10/2009
• Est. Priority Date: 11/17/2004
• Status: Expired due to Fees

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Yoon, Tae-ho, Kim, Kyeong-seop, Kim, Kyung-ho
• Primary Examiner(s): Lau; Tung S
• Assistant Examiner(s): KHUU, HIEN DIEU THI
• Application Number:
  US11/268,599
• Publication Number:
  US 20060106571A1
• Time in Patent Office: 1,463 Days
• Field of Search: 702/79
• US Class Current: 702/79
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06F 2218/00 Aspects of pattern recognition specially adapted for signal processing
• CPC Class Codes: G06V 40/10 Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands

Abstract

A biometric apparatus and a method thereof using bio signals are provided. The apparatus includes an ADC, a periodic signal extractor, a template storing portion, a comparator. The ADC performs sampling of an input bio signal to convert the sampled bio signal into a digital signal. The periodic signal extractor extracts the periodic signals from the digital signal. The template storing portion registers a plurality of users and stores periodic signals for the respective registered users in a form of templates. The comparator computes similarity between the periodic signals outputted from the periodic signal extractor and the respective templates to select a template whose similarity is greatest and recognizes a user that corresponds to the selected template as a user who corresponds to the input bio signal.

Claims

1. A biometric apparatus, comprising:

an ADC (analog-to-digital converter) sampling an input bio signal to convert the sampled bio signal into a digital signal;

a periodic signal extractor for extracting a plurality of periodic signals from the digital signal;

a template storing portion registering a plurality of users and storing periodic signals for the respective registered users in a form of a template; and

a comparator computing similarity by calculating distances between the periodic signals outputted from the periodic signal extractor and the respective templates to select a template whose calculated distance is smallest and recognizing a user who corresponds to the selected template as a user who corresponds to the input bio signal.

2. The apparatus as claimed in claim 1, wherein the periodic signal is expanded to a 35 harmonic Fourier series.

3. The apparatus as claimed in claim 1, wherein the template is one of sample values, Fourier coefficients, or Fourier coefficients spectrums.

4. The apparatus as claimed in claim 1, wherein the input bio signal is an electrocardiogram signal.

5. A biometric apparatus, comprising:

an ADC (analog-to-digital converter) sampling an input bio signal to convert the sampled bio signal into a digital signal;

a periodic signal extractor for extracting a plurality of periodic signals from the digital signal;

a periodic signal analyzer expanding the period signals to Fourier series to obtain Fourier coefficients for the periodic signals;

a template storing portion registering a plurality of users and storing Fourier coefficients of periodic signals for the respective registered users in form of templates; and

a comparator computing similarity by calculating distances between values outputted from the periodic signal analyzer and the respective templates to select a template whose calculated distance is smallest and recognizing a user who corresponds to the selected template as a user who corresponds to the input bio signal.

6. The apparatus as claimed in claim 5, wherein the calculated distance is computed as distances between the values outputted from the periodic signal analyzer and the respective templates, as given by the following equation:

${\mathrm{distance}}_i=\underset{n}{\sum <br><br>}<br><br>\sqrt{{\left({\stackrel{\_}{c}}_i-c_n\right)}^2}$

where, i=a number of the registered users, n=a number of the periodic signals, c_i;

i-th template Fourier coefficient, c_n;

Fourier coefficient of n-th periodic signal.

7. The apparatus as claimed in claim 6, wherein c_i is a value obtained by averaging the Fourier coefficients for a plurality of periodic signals obtained from an i-th registered user.

8. The apparatus as claimed in of claim 5, wherein the periodic signal analyser further obtains Fourier coefficient spectrums by applying Fourier transform to the Fourier coefficients, and the template storing portion stores the Fourier coefficient spectrum for each of the registered users as the template.

9. The apparatus as claimed in claim 8, wherein if the values outputted from the periodic signal analyser are the Fourier coefficient spectrums, the similarity is computed as distances between the Fourier coefficient spectrums and the respective templates, as given by the following equation:

${\mathrm{distance}}_i=\underset{n}{\sum <br><br>}<br><br><br><br>\sqrt{{\left({\stackrel{\_}{F}}_i<br><br>\left(t\right)-F_n<br><br>\left(t\right)\right)}^2}$

where, i;

a number of the registered users, n;

a number of the periodic signals, F_i;

i-th template Fourier coefficient spectrum, F_n;

Fourier coefficient spectrum of n-th periodic signal.

10. The apparatus as claimed in claim 9, wherein F _i is a value obtained by averaging the Fourier coefficient spectrums for a plurality of periodic signals obtained from an i-th registered user.

11. The apparatus as claimed in claim 5, wherein the periodic signal is expanded to a 35 harmonic Fourier series.

12. The apparatus as claimed in claim 5, wherein the template is one of sample values, Fourier coefficients, or Fourier coefficients spectrums.

13. The apparatus as claimed in claim 5, wherein the input bio signal is an electrocardiogram signal.

14. The apparatus as claimed in claim 5, the comparator authenticates the user if the computed distance between template values of a user selected as an authentication target from the template storing portion if in case of 1:

1 authentication.

15. The apparatus as claimed in claim 5, the comparator recognizes the user as a user who corresponds to a template whose distance between values stored in the template storing portion and values outputted from the periodic signal analyzer is smallest.

16. A biometric method, comprising:

sampling an input bio signal to convert the sampled bio signal into a digital signal;

extracting a plurality of periodic signals from the digital signal;

calculating, using at least one processing device, a similarity by calculating distances between the extracted periodic signals and template from a template storing portion to which a plurality of users are registered and periodic signal for the respective registered users as stored as the templates; and

selecting a template whose calculated distance is smallest and recognizing a user who corresponds to the selected template as a user who corresponds to the input bio signal.

17. The method as claimed in claim 16, wherein the extracting of the plurality of periodic signals further comprises obtaining the Fourier coefficients for the plurality of periodic signals, respectively, and the template storing portion stores corresponding Fourier coefficients for each registered user in a form of the templates.

18. The apparatus as claimed in claim 16, wherein the periodic signal is expanded to a 35 harmonic Fourier series.

19. The apparatus as claimed in claim 16, wherein the template is one of sample values, Fourier coefficients, or Fourier coefficients spectrums.

20. The apparatus as claimed in claim 16, wherein the input bio signal is an electrocardiogram signal.

21. A biometric method, comprising:

sampling an input bio signal to convert the sampled bio signal into a digital signal;

extracting a plurality of periodic signals from the digital signal;

calculating, using at least one processing device, a similarity between the extracted periodic signals and template from a template storing portion to which a plurality of users are registered and periodic signal for the respective registered users as stored as the templates; and

selecting a template whose similarity is greatest and recognizing a user who corresponds to the selected template as a user who corresponds to the input bio signal

wherein the similarity is computed as distances between the periodic signals and the respective templates, as given by the following equation;

${\mathrm{distance}}_i=\underset{n}{\sum <br><br>}<br><br><br><br>\sqrt{{\left({\stackrel{\_}{f}}_i<br><br>\left(t\right)-f_n<br><br>\left(t\right)\right)}^2}$

where, i;

a number of the registered users, n;

a number of the periodic signals, f_i;

sample value of i-th template, f_n;

sample value of n-th periodic signal.

22. The method as claimed in claim 21, wherein f_i(t) is a value obtained by averaging a plurality of periodic signals obtained from an i-th registered user.

23. A biometric method, comprising:

sampling an input bio signal to convert the sampled bio signal into a digital signal;

extracting a plurality of periodic signals from the digital signal;

calculating, using at least one processing device, a similarity between the extracted periodic signals and template from a template storing portion to which a plurality of users are registered and periodic signal for the respective registered users as stored as the templates; and

selecting a template whose similarity is greatest and recognizing a user who corresponds to the selected template as a user who corresponds to the input bio signal,

wherein the extracting of the plurality of periodic signals further comprises obtaining the Fourier coefficients for the plurality of periodic signals, respectively, and the template storing portion stores corresponding Fourier coefficients for each registered user in a form of the templates, and a Fourier transform is applied to the Fourier coefficients to obtain Fourier coefficient spectrums, wherein the template storing portion stores corresponding Fourier coefficient spectrums for each registered user in a form of the templates.

24. The method as claimed in claim 23, wherein the similarity is computed as distances between the Fourier coefficient spectrums and the respective templates, as given by the following equation:

${\mathrm{distance}}_i=\underset{n}{\sum <br><br>}<br><br><br><br>\sqrt{{\left({\stackrel{\_}{c}}_i-c_n\right)}^2}$

where, i;

a number of the registered users, n;

a number of the periodic signals, F_i;

i-th template Fourier coefficient spectrum, F_n;

Fourier coefficient spectrum of n-th periodic signal.

25. The method as claimed in claim 24, wherein F _i is a value obtained by averaging the Fourier coefficient spectrums for a plurality of periodic signals obtained from an i-th registered user.

26. A computer-readable recording medium for recording a program executing a recognition method, comprising:

detecting a bio signal containing periodic signals from a user and sampling the bio signal to convert the sampled signal into a digital signal;

extracting the periodic signals from the digital signal;

reading templates from a template storing portion in which a plurality of users are registered and storing periodic signals for the respective registered users in a form of templates and computing similarity by calculating distances between the templates and the periodic signals of the user, respectively; and

recognizing a user that corresponds to a template whose calculated distance is smallest, as the user.