Knuckle profile indentity verification system

• US 5,594,806 A
• Filed: 06/20/1994
• Issued: 01/14/1997
• Est. Priority Date: 06/20/1994
• Status: Expired due to Fees

• Current Assignee: Personal Information & Entry Access Control Incorporated
• Original Assignee: Personal Information & Entry Access Control Incorporated
• Inventors: Colbert, Charles
• Primary Examiner(s): Boudreal, Leo
• Assistant Examiner(s): Tadayon, Bijan
• Application Number:
  US08/267,584
• Time in Patent Office: 939 Days
• Field of Search: 382/115, 382/209, 382/221, 382/222, 382/116, 382/117, 382/118, 382/119, 382/190, 382/192, 382/195, 382/203, 382/217, 382/218, 382/219, 382/220
• US Class Current: 382/115
• CPC Class Codes: G06V 40/11 Hand-related biometrics; Hand pose recognition
• CPC Class Codes: G06V 40/12 Fingerprints or palmprints
• CPC Class Codes: G07C 9/257 electronically G07C9/26 takes precedence

Abstract

A method and apparatus for verifying/identifying a person based on contour of the knuckles of at least one hand, e.g. a digitized waveform unique to each individual. At an Enrollment Station a microcomputer is connected to a video camera (with CCD array). A candidate user grasps and turns a grip handle on the apparatus, thus positioning a fist before a viewing window and activating the camera to scan the fist and generate a contour of the user'"'"'s knuckle profile. User'"'"'s data comprises a knuckle profile of fist(s), an assigned PIN, and optionally, information such as user'"'"'s name, bank ID number, Social Security Number, and access restrictions. User'"'"'s data may be stored as an electronic waveform profile (template) in an ID card, and simultaneously in a master database containing data of all authorized users in a guarded system (factory, retail store, bank customers, computer data files). Similar apparatus is installed at each Guarded Station (a door, turnstile, gate, ATM, etc.). To gain access or transaction approval at a Guarded Station, the user inserts his ID card into a reader to retrieve the template from the card and store it temporarily for use in a digital signal processor; the system may also retrieve the template from master database. As at the Enrollment Station, the user generates his knuckle profile by turning a grip/handle into position before a camera viewing window.

Claims

1. A biometric geometry method for preparing a unique identity characteristic of a potential user based on at least one of the user'"'"'s fists placed in grip configuration by creating and storing as a template a reference profile of the undulating peaks and valleys of a knuckle pattern unique to each user, the method comprising the steps of

a) positioning the user'"'"'s knuckle contours of said one fist,

b) capturing an image of the knuckles,

c) extracting the knuckle profile from the captured image,

d) digitizing the knuckle profile,

e) storing the digitized profile as a template in a storage medium, and

f) sharpening the peaks and valleys of the knuckle profile by convolving the profile using an autocorrelation algorithm, then storing the resulting autocorrelation curve information as the template in an auxiliary database.

2. The method as defined in claim 1, further comprising

f) storing, along with the user'"'"'s template, unique user data, namely the user'"'"'s assigned personal identification number, name, Social Security number, or account number.

3. A method according to claim 1, for verifying a user'"'"'s identity, and then

g) comparing the user'"'"'s knuckle profile with the user'"'"'s previously stored template,

h) determining the degree of congruence between the user'"'"'s knuckle profile and template by calculating a numerical match score,

i) comparing the match score with a predetermined threshold score, and

j) determining whether the user is to be accepted or rejected based on the magnitude of the match score.

4. A method according to claim 3,

wherein step j) is based on whether the match score equals or exceeds the threshold score.

5. The method set forth in claim 3 wherein the templates of all authorized users are stored in a database, and step (g) is performed by riffling through such database, comparing the user'"'"'s knuckle profile with each stored template, and selecting the stored template with the highest match score, and step (i) is performed by comparing the selected template with the threshold score and making an accept decision if the threshold criterion is met.

6. A method according to claim 1, for verifying a user'"'"'s identity for a number of different users, and then

e1) storing the digitized knuckle profiles of all users as templates in a database,

k) searching the database to match the user'"'"'s profile with the user'"'"'s previously stored template,

h1) determining the degree of congruence between the user'"'"'s profile and a template selected in step k) by calculating a numerical match score,

i1) comparing the match score with a predetermined threshold score, and

j1) determining whether the user is to be accepted or rejected based on the magnitude of the match score.

7. A method as defined in claim 1, wherein

step (b) is performed by an electronic video camera and associated edge detector, said camera generating scan lines sweeping along the knuckle and along the first proximal finger joints, said edge detector extracting the user'"'"'s knuckle profile from the illuminated fist image and thus generating the knuckle profile.

8. A method as defined in claim 7, wherein a combination of a CCD video camera, an incandescent lamp, and an IR filter are used to insure that high contrast is achieved between the knuckles and surroundings.

9. A method as defined in claim 1, wherein

step (b) is performed by an electronic video camera and associated edge detector, said camera generating scan lines sweeping along the knuckle and along the back of the hand, said edge detector extracting the user'"'"'s knuckle profile from the illuminated fist image and thus generating the knuckle profile.

10. The method of claim 1, further including adjusting a threshold score to meet the needs of the facility served, thus relaxing either the Type I error rate or the Type II error rate at the expense of the other.

11. The method set forth in claim 1, further including as step a1) detecting a pulse in the fist to distinguish a living fist from a non-living surrogate.

12. The method set forth in claim 1, wherein the knuckle image is scanned from the head-on position viewing the knuckles and proximal joints of the fist.

13. The method set forth in claim 1, wherein the knuckle image is scanned from above the back of the fist.

14. The method of claim 1, including the additional step of d1) decreasing the size of a profile and template by recording only the distances between the knuckle peaks and valleys and storing such modified profile as a template in an auxiliary database.

15. The method of claim 1, further including

c2) extracting a profile of the proximal interphalangeal joints in addition to the knuckle profile,

d2) digitizing the joint profile in addition to the knuckle profile, and

wherein step (e) stores both the digitized knuckle profile and digitized joint profile as a template.

16. An apparatus for verifying the identity of a candidate user based on a knuckle profile of at least one fist, comprising

a face plate having a first viewing window therein of a size to accommodate the area of the knuckles of a closed fist,

a grip handle for the candidate user to grasp so as to bring his fist into position for viewing through said window,

means mounting said grip handle for pivotable motion about an axis to one side of said window, and for holding said grip handle normally upright, whereby grasping and turning said grip handle to a position in front of said window will locate the fist for viewing through said window,

profice capturing means mounted so as to view and record through said window an image of the knuckles of a fist, said camera generating a scan-type output waveform signal,

digitizer means for receiving said output signal and transforming said output signal into a digitized profile of the knuckles, and

means for storing the digitized profile signals in memory as a template.

17. Apparatus as defined in claim 16, further including a second viewing window located in spaced relation to said first window and said grip handle,

said mounting means for said grip handle providing for turning of said grip handle to positions in front of either of said first and second windows,

whereby a user can present his one fist before said first window and his second fist before said second window,

said camera viewing and scanning the images through said windows in sequence and said digitizer means transforming the two images into a combined output signal for storing as a single template.

18. Apparatus as defined in claim 16, further including

a card reader/encoder for storing a user'"'"'s profile waveform signal to be possessed by the user associated therewith.

19. Apparatus as defined in claim 16, further including

means providing a database memory for storing individually as templates the profile signals from the digitizer means, and

means for comparing the stored templates with a digitized profile to determine congruence between the profile generated from a user'"'"'s knuckles and a certain one of the stored templates.