Fingerprint identification method and apparatus

• US 7,194,115 B2
• Filed: 07/26/2001
• Issued: 03/20/2007
• Est. Priority Date: 07/28/2000
• Status: Expired due to Fees

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Uchida, Kaoru
• Primary Examiner(s): Wu; Jingge
• Assistant Examiner(s): CARTER, AARON W
• Application Number:
  US10/333,755
• Publication Number:
  US 20030123715A1
• Time in Patent Office: 2,063 Days
• Field of Search: 382124-127, 382115-123, 340/5.1, 340/5.2, 340/5.52, 340/5.53, 902/3, 713/186
• US Class Current: 382/124
• CPC Class Codes: G06V 40/1335 Combining adjacent partial images (e.g. slices) to create a composite input or reference pattern; Tracking a sweeping finger movement

Abstract

A method and device for fingerprint identification implementing precise identification at a high speed by using partial images of a fingerprint obtained through relative movements of a finger to a small sized sensor are provided. The device includes a frame input unit for inputting a partial image of a fingerprint. A registered image optimum position calculating unit compares the partial image to a registered fingerprint and an image combining unit creates a composite of the partial images using the optimum position. A fingerprint collating unit judges the similarity between the composite image and registered fingerprint to determine the correlation between the partial images and the registered fingerprint.

Claims

1. A fingerprint identification method, in which a plurality of partial images of a fingerprint is inputted and the similarity between the plurality of partial images and a registered fingerprint image is judged, comprising:

determining an optimal position for each of the partial images within the registered fingerprint image in which a first penalty index which represents a difference between one of the plurality of partial images and the registered fingerprint image is minimized for each partial image;

arranging and combining the partial images based the optimal position for each partial image to form a composite image having an area;

accumulating the first penalty index for each partial image to determine a total first penalty value; and

determining that the plurality of partial images differs from the registered fingerprint image when the total first penalty value exceeds a predetermined penalty threshold, when the area of the composite image exceeds a predetermined threshold.

2. A fingerprint identification method, in which a plurality of partial images of a fingerprint is inputted and the similarity between the plurality of partial images and a previously registered fingerprint image is judged, comprising:

determining an optimal position for each of the partial images within the registered fingerprint image in which a first penalty index which represents a difference between one of the plurality of partial images and the registered fingerprint image is minimized for each partial image;

arranging and combining the partial images based the optimal position for each partial image to form a composite image;

collating the composite image with the registered fingerprint image to determine an amount of similarity between the composite image and the registered fingerprint image;

and

determining that the plurality of fingerprint images input differs from the registered fingerprint image when the amount of similarity falls below a predetermined penalty threshold.

3. A fingerprint identification method, in which a plurality of partial images of a fingerprint are inputted and the similarity between the plurality of partial images and a registered fingerprint image is judged, comprising:

determining an optimal position for each of the partial images within the registered fingerprint image in which a first penalty index which represents a difference between one of the plurality of partial images and the registered fingerprint image is minimized for each partial image;

accumulating the first penalty index for each partial image to determine a total first penalty value;

arranging and combining the partial images based the optimal position for each partial image to form a composite image having an area; and

determining that the plurality of partial images resembles the registered fingerprint image when the total first penalty value does not exceed a predetermined penalty threshold, and the area of the composite image becomes larger than a predetermined area threshold.

4. A fingerprint identification method, in which a plurality of partial images of a fingerprint are inputted and the similarity between the plurality of partial images and a registered fingerprint image is judged, comprising:

determining an optimal position for each of the partial images within of the registered fingerprint image in which a first penalty index which represents a difference between one of the plurality of partial images and the registered finger print image is minimized for each partial image;

arranging and combining the partial images based on the optimal position for each partial image to form a composite image having an area;

collating the composite image with the registered fingerprint image to determine an amount of similarity between the composite image and the registered fingerprint image;

and

determining that the plurality of fingerprint images resembles the registered fingerprint image when the amount of similarity does not exceed a predetermined similarity threshold, and the area of the composite image becomes larger than a predetermined area threshold.

5. A fingerprint identification method, in which a plurality of partial images of a fingerprint is inputted and the similarity between the plurality of partial images and a registered fingerprint image is judged, comprising:

determining an optimal position for each of the partial images within the registered fingerprint image in which a first penalty index which represents a difference between one of the plurality of partial images and the registered fingerprint image is minimized for each partial image;

arranging and combining the partial images based on the optimal position for each partial image to form a composite image having an area;

collating the composite image with the registered fingerprint image to determine an amount of similarity between the composite image and the registered fingerprint image;

finding a composite position where one of the plurality of partial image fits with least discordance in the composite image; and

arranging and combining the partial images based on the composite position.

6. The fingerprint identification method claimed in claim 5, further comprising:

calculating a second penalty index of each partial image such that the second penalty index which represents a difference between each one of the partial images and the partial composite image is a minimize; and

wherein the optimal position is determined based on the first penalty index and second penalty index.

7. The fingerprint identification method claimed in claim 6,

wherein the optimal position is determined based on the first penalty index and second penalty index in which effect of the second penalty index in determining the optimal position is increased as each partial images is arranged and combined into the partial composite image.

8. A fingerprint identification device for judging the similarity between fingerprint and a previously registered fingerprint image using a plurality of partial images of the fingerprint, comprising:

a frame image input unit which inputs the partial images of the fingerprint;

a registered image optimum position calculating unit which calculates a first penalty index which represents a difference between one of the partial images and the registered fingerprint image when the one of the partial images is at comparison positions;

an image combining unit which combines and arranges the partial images at a position bearing the closest correlation with a partial composite image composed of previously combined and arranged partial images;

a fingerprint collating unit which judges the similarity between a composite image composed of the partial images and the registered fingerprint image; and

a composite image optimum position calculating unit which determines an optimal position where the partial image fits with least discordance in the partial composite image composed of previously combined and arranged partial images;

wherein

the image combining unit combines and arranges the partial images according to the first penalty index and the optimal position.

9. A fingerprint identification device for judging the similarity between fingerprint and a previously registered fingerprint image using a plurality of partial images of the fingerprint, comprising:

a frame image input unit which input the partial images of the fingerprint;

a registered image optimum position calculating unit which calculates first a penalty index which represents a difference between one of the partial images and the registered fingerprint image;

an image combining unit which combines and arranges the partial images at a position bearing the closest correlation with a partial composite image composed of previously combined and arranged partial images;

a fingerprint collating unit which judges the similarity between a composite image composed of the partial images and the registered fingerprint image; and

a mismatch determination unit which accumulates the first penalty index for each of the inputted partial images and determines that the fingerprint differs from the registered fingerprint image when the accumulated first penalty indices exceeds a predetermined penalty threshold.

10. A fingerprint identification device for judging the similarity between a fingerprint and a previously registered fingerprint image using a plurality of partial images of the fingerprint, comprising:

a frame image input unit which inputs the partial images of the fingerprint;

a registered image optimum position calculating unit which calculates a first penalty index which represents a difference between one of the partial images and the registered fingerprint image when the one of the partial images at a comparision position;

an image combining unit which combines and arranges the partial images at a position bearing the closest correlation with a partial composite image composed of previously combined and arranged partial images;

a fingerprint collating unit which judges the similarity between a composite image composed of the partial images and the registered fingerprint image;

a composite image optimum position calculating unit which determines an optimal position where the partial image fits with least discordance in the partial composite image composed of previous combined and arranged partial images; and

a mismatch determination unit which accumulates the first penalty index for each of the partial images and determines that the fingerprint differs from the registered fingerprint image when a value of the accumulated first penalty indices exceeds a predetermined penalty threshold.

11. A fingerprint identification device for judging the similarity between fingerprint and a previously registered fingerprint image using a plurality of partial images of the fingerprint, comprising:

a frame image input unit which inputs the partial images of the fingerprint;

a registered image optimum position calculating unit which calculates a first penalty index which represents a difference between one of the partial images and the registered fingerprint image;

an image combining unit which combines and arranges the partial images at a position bearing the closest correlation with a previous partial composite image to create a partial composite image;

a fingerprint collating unit which judges the similarity between a composite image composed of the partial images and the registered fingerprint image; and

a rough match determination unit which accumulates each first penalty index, and determines that the fingerprint matches the registered fingerprint image when an accumulated value of first penalty indices does not exceed a predetermined penalty threshold, and the area of the partial composite image combined by the image combining unit exceeds a predetermined area threshold.

12. A fingerprint identification device for judging the similarity between a fingerprint and a previously registered fingerprint image using a plurality of partial images of the fingerprint, comprising:

a frame image input unit which inputs the partial images of the fingerprint;

a registered image optimum position calculating unit which calculates a first penalty index which represents a difference between one of the partial images and the registered fingerprint image;

an image combining unit which combines and arranges the partial image at a position bearing the closest correlation with a previous partial composite image composed of previously combined and arranged partial images to form a partial composite image having an area;

a fingerprint collating unit which judges the similarity between a composite image composed of the partial images and the registered fingerprint image;

a composite image optimum position calculating unit which finds an optimal position where the partial image fits with least discordance in the partial composite image composed of previous combined and arranged partial images; and

a rough match determination unit which accumulates each first penalty index for each partial image, and determines that the fingerprint matches the registered fingerprint image when a value of the accumulated first penalty indices does not exceed a predetermined penalty threshold, and the area of the partial composite image combined by the image combining unit becomes larger than a predetermined area threshold;

wherein

the image combining unit combines and arranges the partial images according to the first penalty index and the optimal position.