Gait waveform feature extracting method and individual identification system

• US 7,346,464 B2
• Filed: 10/28/2003
• Issued: 03/18/2008
• Est. Priority Date: 10/29/2002
• Status: Expired due to Fees

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Takiguchi, Kiyoaki
• Primary Examiner(s): Barlow; John
• Assistant Examiner(s): KUNDU, SUJOY K
• Application Number:
  US10/527,847
• Publication Number:
  US 20060158173A1
• Time in Patent Office: 1,603 Days
• Field of Search: 702/70, 702/75, 702/66, 702/73, 702/150, 702/152, 702/189, 731/72, 324 7612- 7613, 340/658, 340/944, 340/870.26
• US Class Current: 702/70
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06F 2218/12 Classification; Matching
• CPC Class Codes: G06V 40/25 Recognition of walking or running movements, e.g. gait recognition

Abstract

A gait waveform feature extracting method and an individual identification system extract features of the gait waveform. A one-step waveform corresponding to one step of a walking movement is specified using, as an index, a peak amplitude corresponding to a state where substantially a whole bottom surface of one foot is in contact with the ground and a toe of the other foot is just after leaving the ground among the electric field displacement formed on the human body in accordance with the human body'"'"'s walking movements. Based on the specified one-step waveform, the features of the one-step waveform are extracted, so that the peak amplitude appears without influence from electric-charge interference between the right and left legs. Accordingly, the one-step waveform reflects the actual one step of the walking movement, and therefore, the features of the one-step waveform can be precisely extracted.

Claims

1. A gait waveform feature extracting method comprising:

specifying a one-step waveform from a portion of a digital signal, said digital signal corresponding to an electric field displacement formed on a body of a subject in accordance with a two-leg-walking movement of said subject, said specifying including

associating as an index of said one-step waveform a peak amplitude in a predetermined frequency band, said peak amplitude corresponding to a state where approximately a whole bottom surface of a first foot of said subject is in contact with a walking surface and a toe of a second foot of said subject is just after leaving the walking surface;

extracting features of said one-step waveform after said one-step waveform is specified in said specifying step; and

identifying said subject based on said features extracted by said extracting.

2. The gait waveform feature extracting method according to claim 1, wherein said predetermined frequency band is in an inclusive range of 6 Hz through 10 Hz.

3. The gait waveform feature extracting method according to claim 1, further comprising:

retrieving the digital signal from memory.

4. The gait waveform feature extracting method according to claim 1, wherein said identifying includes

comparing said features of said one-step waveform against a second waveform stored in memory;

and

determining that said one-step waveform matches said second waveform when said features of said one-step waveform are within a predetermined criteria of corresponding features of said second waveform.

5. The gait waveform feature extracting method according to claim 4, further comprising:

generating a control signal; and

actuating another device once said determining determines that the one-step waveform matches said second waveform.

6. The gait waveform feature extracting method according to claim 5, wherein said actuating includes at least one of actuating a visual display, actuating an audio alarm, and opening a lock.

7. The gait waveform feature extracting method according to claim 5, wherein said determining includes calculating a Mahalanobis distance from said features of said first waveform.

8. The gait waveform feature extracting method according to claim 1, wherein said extracting step includes dividing said one-step waveform by an interval so as to create divided intervals, and extracting as the features of said one-step waveform integrated values obtained by integrating amplitude values of said divided intervals.

9. The gait waveform feature extracting method according to claim 1, further comprising:

generating said digital signal with an electric field displacement detector.

10. The gait waveform feature extracting method according to claim 9, wherein said generating includes producing said digital signal as a wireless signal.

11. The gait waveform feature extracting method according to claim 10, wherein said extracting is performed in an analysis device that is separate from said electric field displacement detector.

12. An individual identification system comprising:

an electric field displacement detector configured to detect an electric field displacement formed on a body of a subject in accordance with a two-leg-walking movement of said subject and produce a signal that corresponds with the electric field displacement; and

a processor configured to identify from said signal an individual using, as an index, a peak amplitude of said signal, in a predetermined frequency band, that corresponds to a state where approximately a whole bottom surface of a first foot of said subject is in contact with a walking surface and a toe of a second foot is just after leaving the walking surface.

13. The individual identification system according to claim 12, wherein said predetermined frequency band is in an inclusive range of 6 Hz through 10 Hz.

14. The individual identification system according to claim 13, further comprising:

a memory configured to hold features of a second waveform associated with the individual,

wherein said processor is configured to compare said features of said one-step waveform against the second waveform stored in memory, and determine that said one-step waveform matches said second waveform when said features of said one-step waveform are within a predetermined criteria of corresponding features of said second waveform.

15. The individual identification system according to claim 14, wherein said electric field displacement detector is configured to generate a control signal when said processor determines that said one-step waveform matches said second waveform;

and said processor is configured to actuate another device after receiving said control signal once said determining step determines that the one-step waveform matches said second waveform.

16. The individual identification system according to claim 15, wherein said another device being at least one of a visual display, an audio alarm mechanism, and a controllable lock.

17. The individual identification system according to claim 14, wherein said processor is configured to calculate a Mahalanobis distance from said features of said first waveform.

18. The individual identification system according to claim 15, wherein said electric field displacement detector includes a transmitter configured to transmit said control signal as a wireless signal.

19. The individual identification system according to claim 14, wherein said electric field displacement detector is separate from said processor.

20. An individual identification system comprising:

means for detecting an electric field displacement formed on a subject in accordance with a two-leg-walking movement of said subject to generate a waveform; and

means for comparing said waveform with predetermined waveforms associated with different individuals so as to identify a predetermined individual based on said electric field displacement.