SUPERVISED EDGE DETECTION USING FRACTAL SIGNATURES

• US 20090297055A1
• Filed: 05/27/2008
• Published: 12/03/2009
• Est. Priority Date: 05/27/2008
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Panda, Subhankar
• Granted Patent: US 8,396,297 B2
• Time in Patent Office: Days
• US Class Current: 382/249
• CPC Class Codes: G06T 7/13 Edge detection
• CPC Class Codes: G06V 10/443 by matching or filtering

Abstract

Methods and systems to identify image pixels as edge pixels using fractal signatures associated with the image pixels. Fractal signatures may include one or more of a variety of fractal dimensions. A fractal dimension of a pixel may be generated from an array of pixels that include the pixel, from x and y coordinates and one or more of luminosity values and color values associated with pixels in the array of pixels. Pixels may be identified as edge pixels when their corresponding fractal signatures are equal to or greater than a fractal signature threshold. The fractal signature threshold may be generated in a supervised fashion.

Claims

1. A system, comprising:

fractal signature logic to generate fractal signatures of a plurality of pixels of an image;

edge pixel identification logic to identify one or more of the plurality of pixels as edge pixels when a corresponding fractal signature is at least equal to a fractal signature threshold; and

output logic to output an indication of the one or more edge pixels to one or more of a user, a printer, a display, a memory device, and integrated circuit logic.

2. The system of claim 1, wherein the fractal signature logic includes logic to generate fractal signatures of a plurality of training edge pixels, the system further comprising:

threshold logic to generate the fractal signature threshold from the fractal signatures of the training edge pixels.

3. The system of claim 2, wherein one or more of the fractal signature logic, the threshold logic, the edge pixel identification logic, and the output logic comprise a computer program product including a computer readable medium having computer program logic stored therein, the computer program product logic including at least a portion of one or more of:

the fractal signature logic to cause the computer system to generate the fractal signatures of the plurality of pixels and of the training edge pixels;

the threshold logic to cause the computer system to generate the fractal signature threshold from the fractal signatures of the training edge pixels;

the edge pixel identification logic to cause the computer system to identify the one or more of the plurality of pixels as edge pixels when the corresponding fractal signature is at least equal to the fractal signature threshold; and

the output logic to cause the computer system to output the identification of the one or more edge pixels.

4. The system of claim 2, wherein at least a portion of one or more of the fractal signature logic, the threshold logic, the edge pixel identification logic, and the output logic further comprise integrated circuit logic.

5. The system of claim 2, wherein the threshold logic comprises logic to generate the fractal signature threshold as an arithmetic mean of the fractal signatures of the training edge pixels.

6. The system of claim 2, further comprising receive logic to receive a user-indication of the training edge pixels.

7. The system of claim 2, wherein the fractal signature logic includes fractal dimension logic to generate, for each of the plurality of pixels and for each of the training edge pixels, a fractal dimension of an array of pixels that include the corresponding pixel, from x and y coordinates and one or more of luminosity values and color values associated with pixels in the array of pixels.

8. The system of claim 2, wherein the edge pixel identification logic includes logic to compare a first set of the fractal signatures corresponding to a first portion of the image to a first fractal signature threshold, and to compare a second set of the fractal signatures corresponding to a second portion of the image to a second fractal signature threshold.

9. A method, comprising:

generating, in a system including one or more of an integrated circuit system, a computer system, and a combination of an integrated circuit system and a computer system, fractal signatures of a plurality of pixels of an image;

identifying, in the system, one or more of the plurality of pixels as edge pixels when a corresponding fractal signature is at least equal to a fractal signature threshold; and

outputting, from the system, an indication of the one or more edge pixels to one or more of a user, a printer, a display, a memory device, and integrated circuit logic.

10. The method of claim 9, further comprising:

generating, in the system, fractal signatures for a plurality of training edge pixels; and

generating, in the system, the fractal signature threshold from the fractal signatures of the training edge pixels.

11. The method of claim 10, wherein the generating the fractal signature threshold comprises determining an arithmetic mean of the fractal signatures of the training edge pixels.

12. The method of claim 10, further comprising receiving a user-indication of the training edge pixels.

13. The method of claim 10, wherein the generating the fractal signature comprises generating, for each of the plurality of pixels and for each of the training edge pixels, a fractal dimension of an array of pixels that include the corresponding pixel, from x and y coordinates and one or more of luminosity values and color values associated with pixels in the array of pixels.

14. The method of claim 10, wherein the identifying comprises:

comparing a first set of the fractal signatures corresponding to a first portion of the image to a first fractal signature threshold; and

comparing a second set of the fractal signatures corresponding to a second portion of the image to a second fractal signature threshold.

15. A computer program product including a computer readable medium having computer program logic stored therein to cause a computer system to identify edge pixels, the computer program product logic comprising:

fractal signature logic to cause the computer system to generate fractal signatures of a plurality of pixels of an image;

edge pixel identification logic to cause the computer system to identify one or more of the plurality of pixels as edge pixels when a corresponding fractal signature is at least equal to a fractal signature threshold; and

output logic to cause the computer system to output an indication of the one or more edge pixels to one or more of a user, a printer, a display, a memory device, and integrated circuit logic.

16. The computer program product of claim 15, wherein the fractal signature logic includes logic to cause the computer system to generate fractal signatures for a plurality of training edge pixels, the computer program logic further comprising:

threshold logic to cause the computer system to generate the fractal signature threshold from the fractal signatures of the training edge pixels.

17. The computer program product of claim 16, wherein the threshold logic includes logic to cause the computer system to determine an arithmetic mean of the fractal signatures of the training edge pixels.

18. The computer program product of claim 16, further comprising receive logic to cause the computer system to receive a user-indication of the training edge pixels.

19. The computer program product of claim 16, wherein the fractal signature logic includes logic to cause the computer system to generate, for each of the plurality of pixels and for each of the training edge pixels, a fractal dimension of an array of pixels that include the corresponding pixel, from x and y coordinates and one or more of luminosity values and color values associated with pixels in the array of pixels.

20. The computer program product of claim 16, wherein the edge pixel identification logic includes logic to cause the computer system to compare a first set of the fractal signatures corresponding to a first portion of the image to a first fractal signature threshold, and to compare a second set of the fractal signatures corresponding to a second portion of the image to a second fractal signature threshold.