Method and device for recognizing non-unrolled fingerprints

• US 6,198,836 B1
• Filed: 04/27/1998
• Issued: 03/06/2001
• Est. Priority Date: 08/27/1996
• Status: Expired due to Fees

• Current Assignee: A-Tronic MGM AG, KABA Schliesssysteme AG
• Original Assignee: A-Tronic MGM AG
• Inventors: Hauke, Rudolf
• Primary Examiner(s): MEHTA, BHAVESH M
• Application Number:
  US09/066,310
• Time in Patent Office: 1,044 Days
• Field of Search: 382/115, 382/124-127
• US Class Current: 382/125
• CPC Class Codes: G06V 40/13 Sensors therefor

Abstract

For the better identification of not-enrolled fingerprint images, images of skin lines of two juxtaposed fingers (1, 2) of a hand are recorded. The orientations of the finger longitudinal axes (L1, L2) with respect to the normal direction (N) are determined and by means of a guidance means (13) a rotation or twisting (W1) of the fingers about their longitudinal axis is largely avoided, so that virtually coinciding, defined partial image quantities (B3, B4) of the complete enrolled fingerprint (A) are detected and transferred by evaluation algorithms into a reduced data quantity. This allows a much greater identification reliability with a smaller storage requirement and simpler evaluation algorithms.

Claims

1. Method for the identification of not-enrolled fingerprint images, wherein images are recorded of skin lines of at least two juxtaposed fingers (1, 2) of a hand, orientations of finger longitudinal axes (L1, L2) are determined, and a rotation (W1) of the fingers about their longitudinal axis is largely prevented by a guidance means (13), so that during each identification process clearly defined, almost identical partial fingerprint image quantities (B3, B4) of skin lines of the two juxtaposed fingers of whole, enrolled fingerprints are detected, therefrom digital grey value images are determined and transferred by evaluation algorithms into a digital biometric pattern with a reduced data quantity.

2. Method according to claim 1, wherein the finger longitudinal axes (L1, L2) are determined by establishing centroid lines of each fingerprint image and from said longitudinal axes are determined their orientation angles (α

1, α

2) and their aperture angles δ

=α

1−

α

2.

3. Method according to claim 1, wherein the finger longitudinal axes (L1, L2) are kept parallel and in a normal direction N by a parallel guidance means (α

1=α

2=0).

4. Method according to claim 1, wherein core positions (C1, C2) of the at least two fingers (1, 2) are determined and from said core positions a spacing (R) and direction (β

) are established.

5. Method according to claim 1, wherein further biometric parameters, comprising at least one of the following features:

centers of gravity of fingerprints, spacings (D1, D2) and a relative position of the skin folds of the top joints, are determined.

6. Method according to claim 1, wherein an associated data bank with comparison patterns (19) is indexed to primary biometric data.

7. Method according to claim 6, wherein core positions (C), locations, angles and spacing details (D) are used as primary biometric data.

8. Method according to claim 1, wherein restricted image section quantities (S, S3) are used for evaluating with evaluation algorithms (15).

9. Apparatus for the identification of not-enrolled fingerprint images comprising an image recording device (10) for recording of skin lines of at least two juxtaposed fingers (1, 2) of a hand with a support plate (12) on which the two recording fingers (1, 2) are positioned in unturned manner relative to their longitudinal axes (L1, L2) by guidance means (13) and where orientations of finger longitudinal axes (L1, L2) are determined and an evaluating electronics (14) for determining digital grey value images, and with evaluation algorithms (15) for defining and evaluating features of the at least two digital fingerprint images and for reduction into a biometric pattern with a reduced data quantity and an associated data bank (19) with comparison patterns.

10. Apparatus according to claim 9, wherein the guidance means (13) are constituted by a planar support surface for at least four fingers of a hand (31) or for the whole hand (32), whereof a surface part area is constructed as an image recording plate (12) for the two recording fingers (1, 2).

11. Apparatus according to claim 9, wherein a lateral stop face (22) defining the normal direction (N) is placed on the support plate and to which an outside of a recording finger can be applied, parallel to a normal direction.

12. Apparatus according to claim 9, wherein the guidance means is constituted by a register rail (25) in a normal direction (N) and to which on either side the inside of both recording fingers (1,2) can be applied.

13. Apparatus according to claim 12, wherein the fingers (1, 2) each have sides, characterized in that sensors (26) on the register rail (25) on both sides monitor contact with respective sides of the fingers (1, 2).

14. Apparatus according to claim 9, wherein one finger (1, 2) is longer than the other finger (1, 2) and wherein a front stop (28) for the longer finger is provided.

15. Apparatus according to claim 14, wherein the front stop (28) is in contact with a finger pad (7) of the longer finger below a fingernail.

16. Apparatus according to claim 14, wherein on the front stop (28) a clearly defined pressure range for the contact with the finger of 1 to 2N can be set by means of a sensor (29).

17. Method for the identification of not-enrolled fingerprint images, wherein images are recorded of skin lines of at least two juxtaposed fingers (1,2) of a hand, orientations of finger longitudinal axes (L1, L2) are determined by establishing centroid lines of each fingerprint image and from said longitudinal axes are determined their orientation angles (α

1, α

2) and their aperture angle δ

=α

1−

α

2, and a rotation (W1) of the fingers about their longitudinal axis is largely prevented by a guidance means (13), so that during each identification process clearly defined, almost identical partial image quantities (B3, B4) of whole, enrolled fingerprints are detected therefrom digital grey value images are determined and transferred by evaluation algorithms into a digital biometric pattern with a reduced data quantity.

18. Method for the identification of not-enrolled fingerprint images, wherein images are recorded of skin lines of at least two juxtaposed fingers (1,2) of a hand, orientations of finger longitudinal axes (L1, L2) are determined or will be determined and a rotation (W1) of the fingers about their longitudinal axis is largely prevented by a guidance means (13), so that during each identification process clearly defined, almost identical partial inage quantities (B3, B4) of whole, enrolled fingerprints are detected therefrom digital grey value images are determined and transferred by evaluation algorithms into a digital biometric pattern with a reduced data quantity, wherein restricted image section quantities (S, S3) are used for evaluating with evaluation algorithms (15).

19. Method for the identification of not-enrolled fingerprint images, wherein images are recorded of skin lines of at least two juxtaposed fingers (1, 2) of a hand, orientations of finger longitudinal axes (L1, L2) are determined and a rotation (W1) of the fingers about their longitudinal axis is largely prevented by a guidance means (13), so that during each identification process clearly defined, almost identical partial fingerprint image quantities (B3, B4) of skin lines of the two juxtaposed lingers of whole, enrolled fingerprints are detected and transferred by evaluation algorithms into a digital biometric pattern with a reduced data quantity, said pattern including core positions (C1, C2) of the two juxtaposed fingers and a spacing (R) and a direction (β

) of the core positions.