Method and apparatus for automatic extraction of fingerprint cores and tri-radii

• US 4,156,230 A
• Filed: 11/02/1977
• Issued: 05/22/1979
• Est. Priority Date: 11/02/1977
• Status: Expired due to Term

• Current Assignee: Printrak International, Inc. (Motorola Solutions, Inc.)
• Original Assignee: Rockwell International Corporation
• Inventors: Riganati, John P., Capello, Richard D., Brown, Gerald D., Bolton, Norman L., Csiszer, Alexander
• Primary Examiner(s): Boudreau, Leo H.
• Application Number:
  US05/847,987
• Time in Patent Office: 566 Days
• Field of Search: 340/146.3 E, 340/146.3 Q, 340/146.3 AE, 364/819, 364/820, 364/822
• US Class Current: 382/124
• CPC Class Codes: G06V 40/1359 Extracting features related to ridge properties; Determining the fingerprint type, e.g. whorl or loop

Abstract

A method and apparatus for the smoothing and singularity detection of the ridge contour data of a fingerprint image for classification in an automatic fingerprint identification system. A template scanning window is passed electronically over a digital matrix of ridge contour data to generate a set of correlation values corresponding to each contour data element and to a plurality of reference angle vectors. The correlation values are processed for determination of peaks and valleys. Tests are conducted to determine whether the relative values of peaks and valleys and the relative spacing between peaks and valleys satisfy certain threshold criteria for reliable extraction of fingerprint singularity points for classification. The resultant data, representing the number of correlation peaks and the direction of each, defines the location and angular orientation of cores and deltas of a fingerprint.

Claims

1. In an automatic data processing system for classifying contour patterns into preselected classifications, the machine method comprising the steps of

(1) Determining the peaks and valleys in correlation of ridge contour data elements with plurality of reference angle vectors,

(2) Testing said peaks and valleys against a plurality of threshold criteria to ascertain which of said peaks are of interest, and

(3) Identifying the number of peaks of interest and the corresponding reference angle vector associated with each identified peak,

whereby singularity detection, and classification of an image are enhanced.

2. In an automatic system for processing fingerprint patterns which are each characterized by epidermal ridge lines forming a contour pattern classifiable into one of a predetermined number of classification types, the automatic system including means for scanning a fingerprint, means for accepting identifying information corresponding to said scanned fingerprint, means for automatically extracting ridge contour data that may be represented by a matrix of ridge contour data elements from said scanned fingerprint corresponding to said contour pattern, means responsive to said ridge contour data for automatically classifying said contour pattern into one of said classification types and means for automatically storing said identifying information according to said corresponding classification type;

an improved method for automatically classifying fingerprints comprising the steps of;

(1) scanning said ridge contour data,

(2) generating correlation data on a one-for-one basis for each said contour data element, said correlation data being indicative of the relative correlation of the average contour ridge flow in the vicinity of each said element and a plurality of reference angle vectors,

(3) processing said correlation data to identify peaks and valleys of correlation data, and

(4) testing in accordance with a set of reliability criteria the relative values of the peaks and valleys of correlation data, the spacing between peaks of correlation data and the spacing between valleys of correlation data for identifying the number of peaks of correlation data and the reference angle vector to which each such peak corresponds.

3. The method as defined in claim 2 wherein said processing step comprises the steps of:

ascertaining the maximum correlation value corresponding to each said contour data element,

ascertaining the minimum correlation value corresponding to each said contour data element,

ascertaining the difference between the maximum and the minimum correlation values that correspond to each said contour data element; and

wherein said testing step comprises the steps of;

comparing said difference with a predetermined threshold value, and terminating and restarting said processing step if said difference is not greater than said threshold value.

4. The method as defined in claim 2 wherein said processing step comprises the steps of:

ascertaining the value of each correlation peak corresponding to said contour data element; and

wherein said testing step comprises the steps of;

comparing said correlation peak value with a predetermined threshold value, and

identifying each said correlation peak value as a good peak only if said peak value is greater than said threshold value.

5. The method as defined in claim 2 wherein said processing step comprises the steps of:

ascertaining the value of each correlation peak corresponding to each said contour data element,

ascertaining the value of each correlation valley corresponding to each said contour data element, and

ascertaining the difference between each said peak value and each said valley value that correspond to each said contour data element; and

wherein said testing step comprises the steps of;

comparing said difference with a predetermined threshold value, and identifying each said correlation peak as a good peak only if said difference is greater than said threshold value.

6. The method as defined in claim 2 wherein said processing step comprises the steps of:

ascertaining the reference angle vector of each correlation peak corresponding to each said contour data element, and

determining the angular spacing between each pair of reference angle vectors corresponding to respective peaks;

and wherein said testing step comprises the steps of;

comparing said spacing with the predetermined threshold spacing, and

identifying each correlation peak as a good peak only if said spacing is greater than said threshold spacing.

7. The method as defined in claim 2 wherein said processing step comprises the steps of:

ascertaining the reference angle vector of each correlation valley corresponding to each of said contour data element and

determining the angular spacing between each pair of reference angle vectors corresponding to respective valleys;

and wherein said testing step comprises the steps of;

comparing said spacing with the predetermined threshold spacing, and

identifying each said correlation peak as a good peak only if said spacing is greater than said threshold spacing.

8. The method as defined in claim 2 wherein said processing step comprises the steps of:

ascertaining the value of the lowest correlation peak corresponding to each said contour data element; and

wherein said testing step comprises the steps of;

comparing said value of said lowest peak with a predetermined threshold value, and

identifying said lowest peak as a good peak only if said lowest peak value is greater than said threshold value.

9. In an automatic fingerprint pattern processing system responsive to extracted ridge contour data for classifying fingerprint patterns into preselected classification types, the apparatus comprising:

means for determining the peaks and valleys in correlation of ridge contour data elements with a plurality of reference angle vectors,

means for testing said peaks and valleys against a plurality of threshold criteria to ascertain which of said peaks are of interest, and

means for identifying the number of peaks of interest and the corresponding reference angle vector associated with each identified peak,

whereby singularity detection and classification of a fingerprint image are enhanced.

10. In an automatic system for processing fingerprint patterns which are each characterized by epidermal ridge lines forming a contour pattern classifiable into one of a predetermined number of classification types, the automatic system including means for scanning a fingerprint, means for accepting identifying information corresponding to said scanned fingerprint, means for automatically extracting ridge contour data that may be represented by a matrix of ridge contour data elements from said scanned fingerprint corresponding to said contour pattern, means responsive to said ridge contour data for automatically classifying said contour pattern into one of said classification types and means for automatically storing said identifying information according to said corresponding classification type;

the improvement wherein said automatic classifying means comprises;

(1) means for scanning said ridge contour data,

(2) means for generating correlation data on a one-for-one basis for each said contour data element, said correlation data being indicative of the relative correlation of the average contour ridge flow in the vicinity of each said element with a plurality of reference angle vectors,

(3) means for processing said correlation data for identifying peaks and valleys of correlation data, and

(4) means for testing in accordance with a set of reliability criteria the relative values of peaks and valleys of correlation data, for testing the spacing between peaks of correlation data and for testing the spacing between valleys of correlation data for identifying the number of peaks of correlation data and the reference angle vector to which each such peak corresponds.

11. The improvement as defined in claim 10, wherein said means for processing comprises:

means for ascertaining the maximum correlation value corresponding to each said contour data element,

means for ascertaining the minimum correlation value corresponding to each said contour data element,

means for ascertaining the difference between the maximum and minimum correlation values that correspond to each said contour data element; and

wherein said means for testing comprises;

means for comparing said difference with a predetermined threshold value, and

means for resetting said means for processing if said difference is not greater than said threshold value.

12. The improvement as defined in claim 10, wherein said means for processing comprises:

means for ascertaining the value of each correlation peak corresponding to each said contour data element; and

wherein said means for testing comprises;

means for comparing each said correlation peak value with a predetermined threshold value, and

means for identifying each said correlation peak value as a good peak only if said peak value is greater than said threshold value.

13. The improvement as defined in claim 10 wherein said means for processing comprises:

means for ascertaining the value of each correlation peak corresponding to each said contour data element,

means for ascertaining the value of each correlation valley corresponding to each said contour data element, and

means for ascertaining the difference between each said peak value and each said valley value that correspond to each said contour data element; and

wherein said means for testing comprises;

means for comparing said difference with a predetermined threshold-value, and

means for identifying each said correlation peak as a good peak only if said difference is greater than said threshold value.

14. The improvement as defined in claim 10 wherein said means for processing comprises:

means for ascertaining the reference angle vector of each correlation peak corresponding to each said contour data element, and

means for determining the angular spacing between each pair of reference angle vectors corresponding to respective peaks,

and wherein said means for testing comprises;

means for comparing said spacing with a predetermined threshold spacing, and

means for identifying each said correlation peak as a good peak only if said spacing is greater than said threshold spacing.

15. The improvement as defined in claim 10 wherein said means for processing comprises:

means for ascertaining the reference angle vector of each correlation valley corresponding to each said contour data element,

means for ascertaining the reference angle vector of each correlation valley corresponding to each said contour data element, and

means for determining the angular spacing between each pair of reference angle vectors corresponding to respective valleys;

and wherein said means for testing comprises;

means for comparing said spacing with a predetermined threshold spacing, and

means for identifying each said correlation peak as a good peak only if said spacing is greater than said threshold spacing.

16. The improvement as defined in claim 10 wherein said means for processing comprises:

means for ascertaining the value of the lowest correlation peak corresponding to each said contour data element; and

wherein said means for testing comprises;

means for comparing said value of said lowest peak with a predetermined threshold value, and

means for identifying said lowest peak as a good peak only if said lowest peak value is greater than said threshold value.

17. In an automatic system for processing fingerprint patterns which are each characterized by ridge lines forming a contour pattern classifiable into one of a predetermined number of classification types, the automatic system including means for scanning a fingerprint, means for accepting identifying information corresponding to said scanned fingerprint, means for automatically extracting ridge contour data that may be represented by a matrix of ridge contour data elements from said scanned fingerprint corresponding to said contour pattern, means responsive to said ridge contour data for automatically classifying said contour pattern into one of said classification types and means for automatically storing said identifying information according to said corresponding classification type;

the improvement wherein said automatic classifying means comprising;

(1) means for scanning said ridge contour data,

(2) means for generating correlation data on a one-for-one basis for each said contour data element, said correlation data being indicative of the relative correlation of the average contour ridge flow in the vicinity of each said element and a plurality of reference angle vectors,

(3) means for identifying peaks and valleys of correlation data for each said contour data element,

(4) means for ascertaining the maximum correlation value corresponding to each said contour data element,

(5) means for ascertaining the minimum correlation value corresponding to each said contour data element,

(6) means for ascertaining the difference between the maximum and minimum correlation values that correspond to each said contour data element.

(7) means for comparing said max-min difference with a predetermined max-min threshold value,

(8) means for resetting said means for identifying if said max-min difference is not greater than said max-min threshold value,

(9) means for ascertaining the value of each correlation peak corresponding to each said contour data element,

(10) means for comparing each said correlation peak value with a predetermined peak threshold value,

(11) means for identifying said correlation peak value as a good peak only if said peak value is greater than said peak threshold value,

(12) means for ascertaining the value of each correlation valley corresponding to each said contour data element,

(13) means for ascertaining the difference between each said peak value and each said valley value that correspond to each said contour data element,

(14) means for comparing said peak-valley difference with a predetermined peak-valley threshold value,

(15) means for identifying said correlation peak as a good peak only if said peak-valley difference is greater than said peak-valley threshold value,

(16) means for ascertaining the reference angle vector of each correlation peak corresponding to each contour data element,

(17) means for determining the angular spacing between each pair of reference angle vectors corresponding to respective peaks,

(18) means for comparing said peak-to-peak spacing with a predetermined peak-to-peak threshold spacing,

(19) means for identifying each correlation peak as a good peak only if said peak-to-peak spacing is greater than said peak-to-peak threshold spacing,

(20) means for ascertaining the reference angle vector of each correlation valley corresponding to each said contour data element,

(21) means for determining the angular spacing between each pair of reference angle vectors corresponding to respective valleys,

(22) means for comparing said valley-to-valley spacing with a predetermined valley-to-valley threshold spacing, and

(23) means for identifying each said correlation peak is a good peak only if said valley-to-valley spacing is greater than said valley-to-valley threshold spacing.