Systems and methods for biometric identification using the acoustic properties of the ear canal

• US 5,787,187 A
• Filed: 04/01/1996
• Issued: 07/28/1998
• Est. Priority Date: 04/01/1996
• Status: Expired due to Fees

• Current Assignee: Sandia Corporation (Martin Marietta Corporation)
• Original Assignee: Sandia Corporation (Martin Marietta Corporation)
• Inventors: Bouchard, Ann Marie, Osbourn, Gordon Cecil
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): MEHTA, BHAVESH M
• Application Number:
  US08/626,704
• Time in Patent Office: 848 Days
• Field of Search: 382/115-118, 340/825.3, 340/825.31, 340/825.34, 902/3, 902/5, 235/379-382, 364/DIG. 1, 364/DIG. 2, 73/579, 73/585, 73/587, 181/0.5, 181/123, 181/126-135, 705/18, 705/44, 395/2.09, 395/2.4, 395/2.55, 395/2.82, 395/2.83, 381/60, 381/68, 381/68.2, 381/68.3, 367/191, 367/13, 128/746, 128/920, 128/630, 128/647, 128/660.01
• US Class Current: 382/115
• CPC Class Codes: A61B 5/12 Audiometering
• CPC Class Codes: A61B 5/7257 using Fourier transforms
• CPC Class Codes: G06V 40/10 Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
• CPC Class Codes: G07C 9/37 using biometric data, e.g. fingerprints, iris scans or voice recognition
• CPC Class Codes: G10L 17/02 Preprocessing operations, e.g. segment selection; Pattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal components; Feature selection or extraction

Abstract

The present invention teaches systems and methods for verifying or recognizing a person'"'"'s identity based on measurements of the acoustic response of the individual'"'"'s ear canal. The system comprises an acoustic emission device, which emits an acoustic source signal s(t), designated by a computer, into the ear canal of an individual, and an acoustic response detection device, which detects the acoustic response signal f(t). A computer digitizes the response (detected) signal f(t) and stores the data. Computer-implemented algorithms analyze the response signal f(t) to produce ear-canal feature data. The ear-canal feature data obtained during enrollment is stored on the computer, or some other recording medium, to compare the enrollment data with ear-canal feature data produced in a subsequent access attempt, to determine if the individual has previously been enrolled. The system can also be adapted for remote access applications.

Claims

1. A system for verifying/recognizing the identity of an individual by characterizing the acoustic properties of the ear canal as a biometric, said system comprising:

enrollment means for enrolling the individual by characterizing the acoustic properties of an ear canal of the individual; and

verification means for verifying the identity of the individual by determining whether the individual has previously been enrolled during said step of enrolling the individual.

2. The system of claim 1, wherein said enrollment means comprises:

data acquisition means for acquiring reference data from the ear canal of the individual;

extraction means for extracting reference feature vector data; and

storage means for storing the reference feature vector data.

3. The system of claim 2, wherein the data acquisition means comprises:

source means for emitting a source signal s(t) as a function of time t into the ear canal of the individual; and

detection means for detecting a response signal f(t) as a function of time in response to the source signal s(t).

4. The system of claim 3, further comprising:

means for digitizing the response signal f(t); and

sampling means for sampling the response signal f(t) at a high sampling rate.

5. The system of claim 4, further comprising storage means for storing the response signal f(t) once it has been digitized by said digitizing means.

6. The system of claim 2, wherein data acquisition means for acquiring data from the ear canal of the individual comprises capture means for capturing frequency and intensity information contained in the response signal f(t).

7. The system of claim 2, wherein the source signal s(t) emitted from said source means is controlled.

8. The system of claim 2, wherein the source signal s(t) emitted from said source means is broad-band noise.

9. The system of claim 1, wherein said verification means comprises:

data acquisition means for acquiring newly-acquired data from the ear canal of the individual;

extraction means for extracting newly-extracted feature vector data; and

pattern recognition means for recognizing a pattern to determine if the individual has previously been enrolled.

10. The system of claim 9, wherein said pattern recognition means comprises:

comparison means for comparing the newly-extracted feature vector data against the reference feature vector data.

11. The system of claim 9, wherein said pattern recognition means comprises a computer-implemented pattern recognition algorithm.

12. The system of claim 9, further comprising authorization means for authorizing access if the individual has been previously enrolled and for denying access if the individual has not been previously enrolled.

13. The system of claim 9, wherein the data acquisition means for acquiring newly acquired-data comprises:

source means for emitting a source signal s(t) as a function of time t into the ear canal of the individual; and

detection means for detecting a response signal f(t) as a function of time in response to the source signal s(t).

14. The system of claim 13, wherein the source signal s(t) is emitted in a series of pure-frequency tones until a steady state is reached in the response signal f(t) for each frequency in the series.

15. The system of claim 1, wherein said enrollment means comprises:

a hand-held apparatus for emitting a source signal s(t) into the ear canal of the individual; and

a computer adapted to incorporate analog-to-digital/digital-to-analog capability running a computer program that instructs the computer to acquire data from the ear canal of the individual;

extract reference feature vector data; and

store the reference feature vector data.

16. The system of claim 15, wherein said hand-held apparatus comprises:

a source signal emitter for emitting a source signal s(t) into the ear canal of the individual, the ear canal responding with a response signal f(t); and

a signal response detector for detecting the response signal f(t).

17. The system of claim 16, wherein said source signal emitter is a speaker, and wherein said signal response detector is a microphone.

18. The system of claim 16, wherein the computer digitizes and samples the response signal f(t), and wherein the response signal f(t) is stored in the computer after it is digitized.

19. The system of claim 18, wherein the response signal f(t) is analyzed via a computer-implemented method by a computer to extract feature vector data.

20. The system of claim 18, wherein the series of frequency tones is the range of zero kHz to fifty kHz.

21. The system of claim 20, wherein the range is advanced in increments of 100 Hz.

22. The system of claim 15, further comprising calibration means for calibrating the hand-held apparatus.

23. A method for verifying/recognizing the identity of an individual by characterizing the acoustic properties of the ear canal as a biometric, said method comprising the steps of:

enrolling the individual by characterizing the acoustic properties of an ear canal of the individual a first time; and

verifying the identity of the individual by characterizing the acoustic properties of the ear canal of the individual a second time and by determining whether the individual has previously been enrolled during said step of enrolling the individual.

24. The method of claim 23, wherein said enrolling step comprises the steps of:

acquiring reference data from the ear canal of the individual;

extracting reference feature vector data from the reference data; and

storing the reference feature vector data.

25. The method of claim 24, wherein the step of acquiring reference data comprises the steps of:

emitting source signals s₁ (t) as a function of time t into the ear canal of the individual; and

detecting response signals f₁ (t) as a function of time in response to the step of emitting source signals s₁ (t).

26. The method of claim 25, further comprising the steps of:

digitizing the response signals f₁ (t); and

sampling the response signals f₁ (t) at a predetermined sampling rate.

27. The method of claim 26, further comprising the step of storing the response signals f₁ (t) after said step of digitizing the response signals f₁ (t).

28. The method of claim 24, wherein said step of acquiring reference data from the ear canal of the individual comprises the step of capturing frequency and intensity information contained in response signals f(t), the response signals f(t) being generated in response to the source signals s(t) emitted in the ear canal of the individual.

29. The method of claim 23, wherein said verifying step comprises the steps of:

acquiring newly-acquired data from the ear canal of the individual;

extracting newly-extracted feature vector data from the newly-acquired data; and

recognizing a pattern in the newly-extracted feature vector data to determine if the individual has previously been enrolled.

30. The method of claim 29, wherein said step of recognizing a pattern comprises the step of comparing the newly-extracted feature vector data against the reference feature vector data obtained during said step of enrolling an individual.

31. The method of claim 29, wherein said step of recognizing a pattern is performed using a computer-implemented pattern recognition algorithm.

32. The method of claim 29, further comprising the step of authorizing access if the individual has been previously enrolled.

33. The method of claim 29, further comprising the step of denying access if the individual has not been previously enrolled.

34. The method of claim 29, wherein the step of acquiring newly-acquired data comprises the steps of:

emitting source signals s₂ (t) as a function of time t into the ear canal of the individual; and

detecting response signals f₂ (t) as a function of time in response to the step of emitting source signals s₂ (t).

35. The method of claim 23, wherein said step of enrolling an individual uses reference feature vector data to characterize the individual, and wherein said step of verifying the identity of the individual uses newly-acquired feature vector to characterize the individual.

36. The method of claim 35, wherein the reference feature vector data and the newly-acquired feature vector data are extracted from response signals f(t), the response signals f(t) being generated in response to source signals s(t) emitted in the ear canal of the individual, the response signals f(t) being described as a convolution of the source signals s(t) with an effective impulse response function of the ear canal h(t).

37. The method of claim 35, wherein the reference feature vector data and the newly-acquired feature vector data are extracted from a discrete Fast Fourier Transform F(ω

) of response signals f(t), where ω

is natural frequency and t is time, the response signals f(t) being generated in response to the source signals s(t) emitted in the ear canal of the individual.

38. The method of claim 35, wherein the reference feature vector data and the newly-acquired feature vector data are directly extracted from source signals s(t) emitted in the ear canal of the individual.

39. The method of claim 35, wherein the reference feature vector data and the newly-acquired feature vector data are described as a discretized log |H|(ω

)², where is the natural frequency and H(ω

) is the Fourier transform of an effective impulse response function of the ear canal h(t).

40. The method of claim 35, wherein the reference feature vector data and the newly-acquired feature vector data are described by a discretized function H(ω

) as a function of natural frequency ω

, where ω

is the natural frequency and H(ω

) is the Fourier transform of an effective impulse response function of the ear canal h(t).

41. The method of claim 35, wherein the reference feature vector data and the newly-acquired feature vector data are described by a discretized function h(t) as a function of time t, where h(t) is an effective impulse response function of the ear canal.

42. A system for biometric identification, said system comprising means for determining the identification of an individual by analyzing the acoustic properties of an ear canal of the individual, said means comprising a computer adapted to incorporate analog-to-digital and digital-to-analog capability running a computer program that instructs the computer to acquire data from the ear canal of the individual;

extract reference features;

store the reference features in a storage medium, and read the reference features from the storage medium to determine the identity of the individual.

43. A method for biometric identification, said method comprising the steps of:

enrolling an individual to obtain reference data;

transmitting signals s(t) into an ear canal of the individual;

detecting response signals f(t) from the ear canal of the individual in response to the transmitting signals step; and

comparing the response signals f(t) with the previously obtained signals emitted from the ear canal of the individual to verify the identity of the individual.

44. The method of claim 43, wherein said enrolling step comprises the steps of:

acquiring data;

extracting feature data; and

storing the feature data.

45. The method of claim 43, wherein said comparing step comprises the step of reading the reference data from a storage means.

46. The method of claim 43, wherein the response signals represent acoustic properties of the ear canal of the individual.

47. A system for biometric identification, said system comprising:

acquisition means for storing biometric information from an ear canal of an individual;

source means for transmitting source signals into an ear canal of the individual;

recording means for recording response signals emitted from the ear canal of the individual; and

verification means for verifying that the response signals match the biometric information from the ear canal of the individual to verify the identity of the individual.

48. The system of claim 47, wherein said verification means comprises a pattern recognition algorithm.

49. The system of claim 47, wherein said pattern recognition algorithm is computer implemented.

50. A method for verifying/recognizing the identity of an individual by characterizing the acoustic properties of the ear canal as a biometric, said method comprising the steps of:

enrolling the individual by characterizing the acoustic properties of an ear canal of the individual, wherein said enrolling step comprises the steps of;

acquiring reference data from the ear canal of the individual by emitting source signals s₁ (t) as a function of time t into the ear canal of the individual and detecting response signals f₁ (t) as a function of time that are generated in response to the source signals s(t);

extracting reference feature vector data from the reference data; and

storing the reference feature vector data; and

verifying the identity of the individual by determining whether the individual has previously been enrolled during said step of enrolling the individual, wherein said verifying step comprises the steps of;

acquiring newly-acquired data from the ear canal of the individual by emitting source signals s₂ (t) as a function of time t into the ear canal of the individual and detecting response signals f₂ (t) as a function of time that are generated in response to the source signals s₂ (t);

extracting newly-extracted feature vector data from the newly-acquired data; and

recognizing a pattern in the newly-extracted feature vector data to determine if the individual has previously been enrolled.

51. The method of claim 50, wherein said step of recognizing a pattern comprises the step of comparing the reference feature vector data against the newly-acquired feature vector data.

52. A system for characterizing an individual by characterizing the acoustic properties of the ear canal as a biometric, said system comprising:

transmitting means for transmitting a first signal into the ear canal of the individual;

receiving means for receiving a second signal characteristic of an individual in response to the first signal; and

means for characterizing the individual by characterizing the acoustic properties of the ear canal as a biometric based on the second signal.

53. The system of claim 52, wherein said transmitting means and said receiving means is a transceiver.

54. The system of claim 52, wherein said transmitting means is a speaker and wherein said receiving means is a microphone.

55. A system for remote access biometric identification, said system comprising:

enrollment means, located at a first location, for enrolling the individual by characterizing the acoustic properties of an ear canal of the individual and for generating and transmitting source signals s(t) into an ear canal of an individual;

receiving means, located at a second location, for receiving said source signals s(t) and for detecting and transmitting response signals f(t) to the first location, the response signals f('"'"'t) being generated in response to source signals s(t); and

verifying means, located at the first location, for receiving response signals f(t) and for verifying the identity of the individual.

56. The system of claim 55, wherein:

said enrollments means is a computer adapted to incorporate analog-to-digital/digital-to-analog capability and running a computer program that instructs the computer to acquire data from the ear canal of the individual;

extract reference feature vector data; and

store the reference feature vector data;

said receiving means is a transceiver; and

said verifying means is the computer.

57. A system for remote access biometric identification, said system comprising:

source means, located at a first location, for emitting source signals s(t) into an ear canal of an individual, the ear canal generating response signals f(t) in response to source signals s(t);

detecting means, located at the first location, for detecting the response signals f(t);

extracting means, located at the first location, for extracting features from the response signals f(t);

transmitting means, located at the first location, for transmitting the features to a second location; and

comparing means, located at the second location, for receiving the features and for comparing the features against reference features that characterize the individual.