Fingerprint classification via spatial frequency components

• US 6,226,391 B1
• Filed: 07/09/1999
• Issued: 05/01/2001
• Est. Priority Date: 07/24/1997
• Status: Expired due to Term

• Current Assignee: Northrop Grumman Systems Corporation (Northrop Grumman Corporation)
• Original Assignee: Litton Systems Incorporated (Northrop Grumman Corporation)
• Inventors: Dennis, Phillip Wayne, Mills, Stuart A., Dydyk, Robert Barry
• Primary Examiner(s): Mancuso, Joseph
• Assistant Examiner(s): Bali, Vikkram
• Application Number:
  US09/350,807
• Time in Patent Office: 662 Days
• Field of Search: 382/124, 382/125, 382/126, 382/127, 382/115, 382/116, 382/117, 382/118, 382/210, 382/211, 382/272, 382/278, 356/71
• US Class Current: 382/125
• CPC Class Codes: G06V 40/1359 Extracting features related to ridge properties; Determining the fingerprint type, e.g. whorl or loop

Abstract

The present invention is a method and apparatus for automatically placing a first unknown image, such as an unknown fingerprint image, into one of a plurality of categories. The invention includes storing in a library a plurality of value series, each of which series is derived from the frequency representation of an image category. The categorization process and apparatus takes the frequency image of a first unknown pattern to create a first frequency image. The frequency image plane of the first (unknown) frequency image is divided into a plurality of frequency image plane regions. Each of the frequency image plane regions may be an angular segment radiating from the origin of the frequency image plane. A region value is assigned to each of the frequency image plane regions based on the total energy in the frequency image in that region. The region values for the first frequency image are combined to generate a first series of region values. The first series of region values is compared in a comparator with each of the stored value series. The comparator preferably performs a correlation function on the pattern or series of the regional values using the one dimensional frequency transform of the spatial representation of the pattern of series of regional values.

Claims

1. A method of placing a first unknown image into one of a plurality of categories, the method comprising the steps of:

taking the frequency image of a first unknown pattern to create a first frequency image;

dividing the first frequency image into a plurality of frequency image regions;

assigning a region value to each of the frequency image regions based on frequency image energy in that region;

combining the region values for the first frequency image to generate a first set of region values; and

comparing the first set of region values with each of a plurality of value series, each of which value series is representative of the frequency image of an image category.

2. The method of claim 1, wherein the step of comparing the first set of region values with each of a plurality of value series comprises encoding the first set of region values into a spatial image, and correlating that spatial image with spatial images corresponding to each of the plurality of value series.

3. The method of claim 2, wherein the correlating step comprises performing an optical correlation.

4. The method of claim 1, wherein the first image is a first fingerprint image.

5. The method of claim 1, wherein the step of dividing the frequency image comprises dividing the frequency image into fewer than 360 frequency regions.

6. The method of claim 1, wherein each of the frequency regions is larger than a pixel.

7. The method of claim 1, additionally comprising the step of storing the plurality of value series to create a plurality of stored value series, and wherein the step of comparing the first set of region values with a plurality of value series, each of which value series is representative of the frequency image of an image category comprises comparing the first set of region values with each of the stored value series.

8. A method of classifying a fingerprint pattern into one of a plurality of categories, the method comprising:

taking a two dimensional frequency representation of the fingerprint pattern;

generating a series of region values for the frequency representation, the generating step comprising the steps of;

dividing the frequency representation into fewer than 360 regions,

assigning a region value to each of the regions of the frequency representation; and

combining the region values to generate a series of region values;

converting the series of region values into a spatial representation thereof; and

correlating the spatial representation with a plurality of stored spatial representations, the stored representations being related to a representative fingerprint pattern for one of the plurality of categories.

9. The method of claim 8, wherein the step of comparing the generated spatial representation with a plurality of stored spatial representations comprises correlating the generated spatial representation with a plurality of stored spatial representations.

10. The method of claim 9, wherein the step of correlating the generated spatial representation with a plurality of stored spatial representations comprises optically correlating the generated spatial representation with a plurality of stored spatial representations.

11. A method of placing a first unknown fingerprint into one of a plurality of categories, the method comprising:

creating a plurality of representations of fingerprints, each of which is representative of a fingerprint category, wherein the step of creating comprises, for each of a plurality of representative fingerprints, the steps of;

taking a two dimensional frequency image of a representative fingerprint for each fingerprint category;

dividing of the frequency image into a plurality of regions;

assigning a region value to each of the image regions based on energy present in the frequency image in that region; and

combining the region values to generate a fingerprint value series;

taking a two dimensional frequency image of a first unknown fingerprint to create a first unknown frequency image;

dividing the first unknown frequency image into a plurality of unknown image regions;

assigning a region value to each of the unknown image regions based on energy present in the first unknown frequency image in that region;

combining the region values for the first unknown frequency image to generate a first unknown fingerprint value series; and

comparing the first unknown fingerprint value series with each of the fingerprint value series representative of the different fingerprint categories.

12. The method of claim 11, wherein the step of comparing the first unknown fingerprint value series with each of the fingerprint value series representative of the different fingerprint categories that represent the different categories comprises convolving the one dimensional Frequency image of the first unknown fingerprint value series with the one dimensional Frequency image of each of the fingerprint value series representative of the different fingerprint categories.

13. The method of claim 11, wherein:

the step of creating a plurality of representations of fingerprints additionally comprises, for each of the plurality of representative fingerprints, the step of generating a representation of the fingerprint value series;

the method additionally comprises the step of generating a representation of the first unknown fingerprint value series; and

the step of comparing the first unknown fingerprint value series with each of the fingerprint value series representative of the different fingerprint categories comprises correlating the representation of the first unknown fingerprint value series with each of the representations of the fingerprint value series representative of the different fingerprint categories.

14. The method of claim 11, wherein the step of comparing the generated representation with each of the stored representations representative of the different fingerprint categories comprises optically comparing the generated spatial representation with a plurality of stored spatial representations.

15. An apparatus for placing a first unknown image into one of a plurality of categories, the apparatus comprising:

a library for storing a plurality of value series, each of which series is representative of the frequency image of an image category;

frequency image apparatus for taking the frequency image of a first unknown pattern to create a first frequency image;

sampling apparatus for dividing the first frequency image into a plurality of frequency image regions, wherein the sampling apparatus additionally assigns a region value to each of the frequency image regions based on energy present in that region of the frequency image, and combines the region values for the first frequency image to generate a first series of region values; and

a comparator for comparing the first series of region values with each of the stored value series.

16. The apparatus of claim 15, wherein the sampling apparatus divides the frequency image into fewer than 360 frequency image regions.

17. The apparatus of claim 15, wherein the comparator comprises a correlator for:

generating a spatial representation of the first series of region values; and

correlating the generated spatial representation with a spatial representation of one or more value series stored in the library.