Automatic fingerprint identification system including processes and apparatus for matching fingerprints

• US 4,790,564 A
• Filed: 02/20/1987
• Issued: 12/13/1988
• Est. Priority Date: 02/20/1987
• Status: Expired due to Term

• Current Assignee: Sagem SA (Safran SA)
• Original Assignee: Morpho Systemes
• Inventors: Didier, Bernard, Vassy, Daniel, Larcher, Philippe, Longepierre, Patrick, Lenci, Michel, Irigoin-Guichandut, Francois
• Primary Examiner(s): Kazenske, E. R.
• Assistant Examiner(s): HEYRANA, PAUL
• Application Number:
  US07/016,884
• Time in Patent Office: 662 Days
• Field of Search: 283/67, 283/68, 283/69, 283/70, 283/74, 354/105, 356/71, 382/4, 382/2, 382/5, 382/48, 382/30, 351/71
• US Class Current: 283/69
• CPC Class Codes: G06V 40/12 Fingerprints or palmprints
• CPC Class Codes: G07C 9/26 using a biometric sensor integrated in the pass
• CPC Class Codes: G07C 9/37 using biometric data, e.g. fingerprints, iris scans or voice recognition

Abstract

The invention is an automatic method and related apparatus for identifying fingerprints by means of comparing the minutiae of each fingerprint in a data base of fingerprints with selected ones of precomputed vector images of search minutiae in a search print to be identified, in order to determine the existence or not of a fit between the minutiae of a file print and the preselected search minutiae images, in position and angle, a result of such comparison being a matching score indicating the probability of a match between the search print and one or more file prints, the basis for selection of such precomputed vector images being the identity between the angle of a file print minutiae and the angle of a precomputed vector images of the search minutiae.

Claims

1. In a method for the automatic identification of fingerprints in which minutiae of a search print to be identified are matched with respect to their respective coordinates of location and angle against the coordinates of location and angle of minutiae of a plurality of file prints in a data base of fingerprints, in order to determine the existence or not of a match between said search print and one or more of said file prints, the improvement comprising the steps of

(1) replicating at least one search print minutia by varying at least one of its coordinates of location and angle, thereby to obtain at least one additional minutia which is different from said search print minutia in at least one of said coordinates and

(2) including said set of search print minutiae to be compared against the minutiae of said file prints.

2. The method of claim 1 in which the step of replicating said minutia is characterized by varying two coordinates of location.

3. The method of claim 1 in which the step of replicating said minutia is characterized by varying the coordinate of angle.

4. The method of claim 1 in which the step of replicating said minutia is characterized by varying each of its coordinates of location and angle.

5. The method of claims 1, 2, 3, or 4 in which the step of replicating is carried out a plurality of times with respect to at least one of the minutiae of said search print in order to generate a plurality of minutiae.

6. In a method for the automatic identification of fingerprints in which minutiae of a search print to be identified are matched with respect to their respective coordinates of location and angle against the coordinates of location and angle of minutiae of a plurality of file prints in a database of fingerprints, in order to obtain a matching score indicative of the degree of matching between said search print and one or more of said file prints, the improvement comprising the steps of

(1) separately for a plurality of minutiae in said search print, computing the respective coordinates of location and angle throughout a range of pre-rotation angles, thereby to obtain a plurality of vector images of each of said search print minutiae, and

(2) separately comparing a plurality of minutiae of a preselected number of file prints in said plurality of file prints, only against those ones of said plurality of vector images having an angle, after pre-rotation, which is equal t the angle of said file minutia being compared.

7. The method of claim 6 in which said range of pre-rotation angles is 360 degrees.

8. The method of claim 6 in which said range of pre-rotation angles is symmetrical with respect to the origin of the coordinate system of said search print and is twice as large as the assumed difference in orientation between the coordinate system of said search print and the coordinate system of sad file prints.

9. The method of claims 6, 7, or 8, in which said range is divided into equal increments.

10. In a method for the automatic identification of fingerprints in which minutiae of a search print to be identified are matched with respect to their respective coordinates of location and angle against the coordinates of location and angle of minutiae of each of a plurality of file prints in a database of fingerprints, in order to obtain a matching score indicative of the degree of matching between said search print and one or more of said file prints, the improvement comprising the steps of

(1) separately for at least one minutia in said search print, identifying whether it is a ridge ending or a bifurcation;

(2) separately for at least one minutia in each of said file prints, identifying whether it is a ridge ending or a bifurcation; and

(3) for each file print compared against said search print, incrementing said matching score thereby obtained by a predetermined amount related to whether respective pairs of file print and search print minutiae being compared are identified as both ridge endings or bifurcations.

18. The method of claims 10 or 11 further comprising the steps of

(1) assigning to at least one of the minutiae of a file print or the search print to which it is to be compared a value indicating the quality of that minutia and

(2) for at least one comparison of a file print minutia against a search print incrementing said matching score thereby obtained by a predetermined amount when either minuitae has a quality factor of a preselected value.

21. The method of claims 10 or 11 further comprising the steps of

(1) assigning to at least one of the minutiae of a file print or the search print to which it is to be compared a value indicating the quality of that minutia and

(2) for at least one comparison of a file print thereby obtained by a predetermined amount, which depends on both the quality of the search minutia and the quality of the file minutia.

11. In a method for the automatic identification of fingerprints in which minutiae of a search print to be identified are matched with respect to their respective coordinates of location and angle against the coordinates of location and angle of minutiae of a plurality of file prints in a database of fingerprints, in order to obtain a matching score indicative of the degree of matching between said search print and one or more of said file prints, the improvement comprising the steps of

12. (1) separately for a plurality of minutiae in said search print, computing the respective coordinates of location and angle throughout a range of pre-rotation angles, thereby to obtain a plurality of vector images of each of said search print minutiae,

(2) separately for at least one minutia in said search print, identifying whether it is a ridge ending or a bifurcation;

(3) separately for at least one minutia in each of said file prints, identifying whether it is a ridge ending or a bifurcation; and

(4) separately comparing a plurality of minutiae of a preselected number of file prints in said plurality of file prints, only against those ones of said plurality of vector images having an angle, after pre-rotation, which is equal to the angle of said file minutia being compared;

(5) for each file print compared against said search print, incrementing said matching score thereby obtained by a predetermined amount related to whether respective pairs of file print and search print minutiae being compared are identified as both ridge endings or bifurcations.

14. The method of claim 12 in which said vector images of all of said search minutiae are arranged in a table in random access memory means, according to angle after pre-rotation.

13. In a method for the automatic identification of fingerprints in which minutiae of a search print to be identified are matched with respect to their respective coordinates of location and angle against the coordinates of location and angle of minutiae of a plurality of file prints in a database of fingerprints, in order to obtain a matching score indicative of the degree of matching between said search print and one or more of said file prints, the improvement comprising the steps of

(1) separately for a plurality of minutiae in said search print, computing the respective coordinates of location and angle throughout a range of pre-rotation angles, thereby to obtain a plurality of vector images of each of said search print minutiae,

(2) collecting all vector images for all search minutiae having the same angle after pre-rotation;

and

(3) separately comparing a plurality of minutiae of a preselected number of file prints in said plurality of file prints, only against those ones of said plurality of vector images having an angle, after pre-rotation, which is equal to the angle of said file minutia being compared.

15. In an apparatus for the automatic identification of fingerprints in which minutiae of a search print to be identified are matched with respect to their respective coordinates of location and angle against the coordinates of location and angle of minutiae of a plurality of file prints in a database of fingerprints, in order to obtain a matching score indicative of the degree of matching between said search print and one or more of said file prints, the improvement comprising the combination of

(1) means for computing the respective coordinates of location and angle, throughout a range of prerotation angles, for a plurality of minutiae in said search print, thereby to obtain a plurality of vector images of each of said search print minutiae,

(2) random access memory means for storing said plurality of vector images;

(3) means for comparing a plurality of minutiae of a preselected number of file prints in said plurality of file prints, only against those ones of said plurality of vector images having an angle, after pre-rotation, which is equal to the angle of said file minutia being compared; and

(4) means for deriving a matching score indicative of the degree of matching between said search print and one or more of said preselected file prints, based upon the comparisons performed by said comparing means.

16. In a method for the automatic identification of fingerprints in which minutiae of a search print to be identified are matched with respect to their respective coordinates of location and angle against the coordinates of the location and angle of minutiae of a plurality of file prints in a database of fingerprints, in order to obtain a matching score indicative of the degree of matching between said search print and one or more of said file prints, the improvement comprising the steps of

(1) assigning to at least one of the minutiae of a file print or of the search print to which it is to be compared a value indicating the quality of that minutia and

(2) for at least one comparison of a file print minutia against a search print minutia, incrementing said matching score thereby obtained by a predetermined amount when either minuitae has a quality factor of a preselected value.

17. In a method for the automatic identification of fingerprints in which minutiae of a search print to be identified are matched with respect to their respective coordinates of location and angle against the coordinates of location and angle of minutiae of a plurality of file prints in a database of fingerprints, in order to obtain a matching score indicative of the degree of matching between said search print and one or more of said file prints, the improvement comprising the steps of

(1) separately for a plurality of minutiae in said search print, computing the respective coordinates of location and angle throughout a range of pre-rotation angles, thereby to obtain a plurality of vector images of each of said search print minutiae,

(2) assigning to at least one of the minutiae of a file print or of the search print to which it is to be compared a value indicating the quality of that minutia;

(3) separately comparing a plurality of minutiae of a preselected number of file prints in said plurality of file prints, only against those ones of said plurality of vector images having an angle, after pre-rotation, which is equal to the angle of said file minutia being compared;

(4) for each file print compared against said search print, incrementing said matching score thereby obtained by a predetermined amount related to whether either of the file print or search print minutiae being compared has a quality factor of a preselected value.

19. In a method for the automatic identification of fingerprints in which minutiae of a search print to be identified are matched with respect to their respective coordinates of location and angle against the coordinates of the location and angle of minutiae of a plurality of file prints in a database of fingerprints, in order to obtain a matching score indiciative of the degree of matching between said search print and one or more of said file prints, the improvement comprising the steps of

(1) assigning to at least one of the minutiae of a file print or of the search print to which it is to be compared a value indicating the quality of that minutia and

(2) for at least one comparison of a file print minutia against a search print minutia, incrementing said matching score thereby obtained by a predetermined amount, which depends on both the quality of the search minutia and the quality of the file minutia.

20. In a method for the automatic identification fingerprints in which minutiae of a search print to be identified are matched with respect to their respective coordinates of location and angle against the coordinates of location and angle of minutiae of a plurality of file prints in a database of fingerprints, in order to obtain a matching score indicative of the degree of matching between said search print and one or more of said file prints, the improvement comprising the steps of

(1) separately for a plurality of minutiae in said search print, computing the respective coordinates of location and angle throughout a range of pre-rotation angles, thereby to obtain a plurality of vector images of each of said search print minutiae,

(2) assigning to at least one of the minutiae of a file print or of the search print to which it is to be compared a value indicating the quality of that minutia;

(3) separately comparing a plurality of minutiae of a preselected number of file prints in said plurality of file prints, only against those ones of said plurality of vector images having an angle, after pre-rotation, which is equal to the angle of said file minutia being compared;

(4) for each file print compared against said search print, incrementing said matching score thereby obtained by a predetermined amount, which depends on both the quality of the search minutia and the quality of the file minutia.