Apparatus and method for matching image characteristics such as fingerprint minutiae

• US 4,896,363 A
• Filed: 04/24/1989
• Issued: 01/23/1990
• Est. Priority Date: 05/28/1987
• Status: Expired due to Term

• Current Assignee: Thumbscan, Inc. (OneSpan, Inc.), Vasco Corp. (OneSpan, Inc.)
• Original Assignee: Thumbscan, Inc. (OneSpan, Inc.)
• Inventors: Glickman, Jeff B., Taylor, Kenneth E.
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Mancuso, Joseph
• Application Number:
  US07/342,666
• Time in Patent Office: 274 Days
• Field of Search: 382/4, 382/5, 382/23, 382/17, 382/26, 356/71
• US Class Current: 382/125
• CPC Class Codes: G06V 40/12 Fingerprints or palmprints
• CPC Class Codes: G07C 9/37 using biometric data, e.g. fingerprints, iris scans or voice recognition

Abstract

A system for matching images in which characteristic points of an image to be tested for a match, such as a fingerprint, are compared with characteristic points of a master image by attempting to match the distances between pairs of master characteristic points with distances between pairs of live characteristic points, whereby the coordinate system of the test image is not required to be aligned with the coordinate system of the master image. The matching system can be implemented in an identification mode in which the live image is attempted to be matched with each of a number of master images, or a verification mode in which the live image is attempted to be matched with a master image that is purported to be the same as the live image.

Claims

1. A method of matching a live image and a master image having any random or nonrandom distribution of characteristic features throughout an image, comprising the steps of:

generating a set of points representative of the characteristic features of the live image;

generating a set of points representative of the characteristic features of the master image; and

evaluating the match between the live characteristic points and the master characteristic points, said step of evaluating consisting essentially of the further steps of;

deriving the distances between a plurality of pairs of the live characteristic points;

deriving the distances between a plurality of pairs of the master characteristic points; and

comparing the live distances with the master distances to determine whether of not the live image matches the master image;

wherein the steps of deriving the distances between pairs of the live characteristic points and deriving the distances between pairs of the master characteristic points comprise for each such step the step of forming, for each point in the set of points, a spectrum of values representing the distances between the point and each other point in the set of points.

2. A method according to claim 1 further comprising the initial steps of:

providing a set of master images; and

receiving an identification of one of the set of master images for matching with the live image.

3. A method according to claim 1 wherein the step of comparing comprises comparing each of the master distance spectra to live distance spectra by determining, for each such comparison of a master distance spectrum to alive distance spectrum, the distance values in the master distance spectrum that match separate distance values in the live distance values in the live distance spectrum within a predetermined tolerance.

4. A method according to claim 3 wherein the step of comparing further comprises indicating a match between the live image and the master image if at least a predetermined proportion of master distance spectra are matched with separate live distance spectra, where a master distance spectrum is deemed to match a live distance spectrum if the proportion of the distance values in the master distance spectrum that match distance values in the live distance spectrum exceeds a predetermined value.

5. A method of identifying a person'"'"'s identity consisting essentially of the steps of:

receiving an identification of selected live minutiae appearing in at least a portion of a selected fingerprint of the person, where the minutiae are identified by their spatial coordinates relative to a reference coordinate system and may be randomly or nonrandomly distributed throughout said fingerprint portion;

deriving the distances between a plurality of pairs of the live minutiae;

providing values for the distances between a plurality of pairs of selected minutiae appearing inn at least a portion of a master fingerprint wherein the minutiae may be randomly or nonrandomly distributed throughout said fingerprint portion; and

comparing the live distance with the master distances to determine whether or not the live fingerprint matches the master fingerprint;

wherein the steps of deriving the distances between pairs of live minutiae and providing the distances between pairs of master minutiae comprise for each such step the step of forming, for each minutia, a spectrum of values representing the distances between the minutia and each other minutia in its set of live or master minutiae.

6. A method according to claim 5 further comprising the steps of:

providing a collection of sets of master minutiae appearing in at least a portion of their associated master fingerprints; and

receiving an identification of one of the sets of master minutiae and deriving the distances between a plurality of pairs of minutiae in that set of master minutiae.

7. A method according to claim 5 wherein the step of comparing further comprises indicating a match for the live fingerprint if at least a predetermined proportion of master distance spectra are matched with live distance spectra, where a master distance spectrum is deemed to match a live distance spectrum if a predetermined proportion of the distance values in the master distance spectrum match distance values in the live distance spectrum within a predetermined tolerance.

8. A method according to claim 5 further comprising the step of generating an indication of a match if the live fingerprint is successfully matched with the master fingerprint, or alternatively generating an indication of the absence of a match if the live fingerprint is unsuccessfully matched with the master fingerprint.

9. An apparatus for matching a live image and a master image having any random distribution of characteristic features throughout an image, comprising:

a first means for receiving and storing a set of points representative of the characteristic features of the live image;

a second means for receiving and storing a set of points representative of the characteristic features of the master image;

computing means for receiving the live and the master characteristic points from the first and second means, deriving a distances between a plurality of pairs of the live characteristic points and deriving the distances between a plurality of pairs of the master characteristic points, wherein the computing means derives, for each characteristic point in each of the sets of live and master characteristic points, a spectrum of values representing the distances between the characteristic point and each other characteristic point in its set of characteristic points; and

means for comparing the live distances with the master distances and initiating a predetermined activity essentially only on the basis of a match between the live distances and the master distances within a predetermined tolerance.

10. An apparatus according to claim 9 wherein the first means further comprises means for sensing the live image and forming a two-tone digital representation thereof;

means for storing the digital representation; and

means for processing the digital representation to generate a set of points representative of characteristic features of the live image.

11. An apparatus according to claim 9, further comprising a third means associated with the first means for receiving an indication of the purported identity of the live image;

and

storage means for storing a plurality of sets of points representative of characteristic features of a plurality of corresponding master images;

wherein the computing means is adapted to receive from the third means a signal identifying the purported identity of the live image and obtain from the storage means a set of points for the master image associated with the purported identity.