Method and apparatus for providing a binary fingerprint image

• US 20030053666A1
• Filed: 09/13/2001
• Published: 03/20/2003
• Est. Priority Date: 09/13/2001
• Status: Active Grant

• Current Assignee: Indian Statistical Institute, Intel Corporation
• Original Assignee: Indian Statistical Institute
• Inventors: Kundu, Malay K., Bhowmick, Partha, Bishnu, Arijit, Acharya, Tinku, Bhattacharya, Bhargab B., Murthy, Chivukula Anjaneya, Dey, Jayanta K.
• Granted Patent: US 7,136,515 B2
• Time in Patent Office: Days
• US Class Current: 382/125
• CPC Class Codes: G06V 40/1347 Preprocessing; Feature extraction

Abstract

Embodiments of a method and/or apparatus to provide a binary image are disclosed.

Claims

1. A method of producing a binary fingerprint image from a gray-scale image, said method comprising:

classifying pixels based, at least in part, on, for a particular pixel, the gray-level topological relationship with neighboring pixels.

2. The method of claim 1, wherein classifying pixels based, at least in part, on, for a particular pixel, the gray-level topological relationship with neighboring pixels comprises:

classifying pixels based, at least in part, on, for a particular pixel, the gradient of the gray-level with at least some immediately adjacent pixels.

3. The method of claim 2, wherein the pixels are classified into one of multiple classes, based, at least in part, on, for a particular pixel, the gradient of the gray-level with at least some immediately adjacent pixels.

4. The method of claim 3, wherein the pixels are classified into one of at least four classes, based, at least in part, on for a particular pixel, the gradient of the gray-level with all immediately adjacent pixels.

5. The method of claim 3, wherein three of the classes comprise:

crest, plateau and valley.

6. The method of claim 3, and further comprising:

for pixels not able to be unambiguously classified based, at least in part, on, for a particular pixel, at least some immediately adjacent pixels, classifying the pixel based, at least in part, on its gray-level value and the relationship of its gray-level value to the average gray-level value of the pixels that have been classified.

7. The method of claim 6, and further comprising:

thinning any resulting ridge lines to be one pixel thick.

8. The method of claim 1, wherein said gray-scale image comprises a gray-scale fingerprint image

9. An article comprising:

a storage medium, said storage medium having stored thereon instructions, that, when executed, result in performance of a method of producing a binary image from a gray-scale image, said method comprising;

classifying pixels based, at least in part, on, for a particular pixel, the gray-level topological relationship with neighboring pixels.

10. The article of claim 9, wherein said instructions, when executed, result in classifying pixels based, at least in part, on, for a particular pixel, the gray-level topological relationship with neighboring pixels in a manner that further comprises classifying pixels based, at least in part, on, for a particular pixel, the gradient of the gray-level with at least some immediately adjacent pixels.

11. The article of claim 10, wherein said instructions, when executed, result in classifying pixels based, at least in part, on, for a particular pixel, the gradient of the gray-level with at least some immediately adjacent pixels in a manner wherein the pixels are classified into one of multiple classes, based, at least in part, on, for a particular pixel, the gradient of the gray-level with at least some immediately adjacent pixels.

12. The article of claim 11, wherein said instructions, when executed, result in classifying pixels in a manner wherein the pixels classified into one of multiple classes are classified into one of at least four classes, based, at least in part, on for a particular pixel, the gradient of the gray-level with all immediately adjacent pixels.

13. The article of claim 11, wherein said instructions, when executed, further result in, for pixels not able to be unambiguously classified based, at least in part, on, for a particular pixel, at least some immediately adjacent pixels, classifying the pixel based, at least in part, on its gray-level value and the relationship of its gray-level value to the average gray-level value of the pixels that have been classified.

14. The article of claim 13, wherein said instructions, when executed, further result in thinning any resulting ridge lines to be one pixel thick.

15. The article of claim 9, wherein said instructions, when executed, result in performance of a method of producing a binary image from a gray-scale image,

17. The apparatus of claim 14, wherein said gray-scale image comprises a gray-scale fingerprint image.

16. An apparatus comprising:

an integrated circuit;

said integrated circuit being adapted to classify pixels of a gray-scale image based, at least in part, on, for a particular pixel, the gray-level topological relationship with neighboring pixels.

18. The apparatus of claim 16, wherein said integrated circuit is further adapted to classify pixels in a manner wherein the pixels are classified into one of at least four classes, based, at least in part, on for a particular pixel, the gradient of the gray-level with all immediately adjacent pixels.

19. The apparatus of claim 18, wherein said integrate circuit is further adapted to, for pixels not able to be unambiguously classified based, at least in part, on, for a particular pixel, all immediately adjacent pixels, classify the pixel based, at least in part, on its gray-level value and the relationship of its gray-level value to the average gray-level value of the pixels that have been classified.

20. The apparatus of claim 19, wherein said integrated circuit is further adapted to thin any resulting ridge lines to be one pixel thick.