Automatic detection and mapping of symmetries in an image

• US 8,494,297 B2
• Filed: 04/05/2007
• Issued: 07/23/2013
• Est. Priority Date: 04/05/2007
• Status: Expired due to Fees

• Current Assignee: Adobe Systems Incorporated (Adobe Inc.)
• Original Assignee: Adobe Systems Incorporated (Adobe Inc.)
• Inventors: Zhang, Hui, Brandt, Jonathan
• Primary Examiner(s): Wang, Claire X
• Assistant Examiner(s): Thirugnanam, Gandhi
• Application Number:
  US11/697,189
• Publication Number:
  US 20080247660A1
• Time in Patent Office: 2,301 Days
• Field of Search: 382/100, 382/115, 382/118, 382/128, 382/173, 382/276, 382/132, 382/186, 382/196, 382/254
• US Class Current: 382/254
• CPC Class Codes: G06V 10/42 Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation

Abstract

A method, system, and computer-readable storage medium are disclosed for determining one or more symmetries in an image comprising a plurality of pixels. A symmetry value may be automatically determined for each of the plurality of pixels. The symmetry value may indicate the strength of one or more symmetries in the image for the respective pixel. The symmetry value may be stored for each of the pixels.

Claims

1. A computer-implemented method for determining one or more symmetries in an image comprising a plurality of pixels, comprising:

performing by a computer;

automatically determining a symmetry value for each of the plurality of pixels in the image, wherein the symmetry value indicates a strength of one or more symmetries in the image for the respective pixel with respect to one or more other pixels in the image, and wherein automatically determining the symmetry value for each of the plurality of pixels in the image comprises;

determining a symmetric pixel for an axis of symmetry for each of the plurality of pixels; and

determining a respective difference between an area around each of the plurality of pixels and an area around the corresponding symmetric pixel, wherein the area around each of the plurality of pixels includes at least the respective pixel, and wherein the area around the corresponding symmetric pixel includes at least the corresponding symmetric pixel; and

storing the symmetry value for each of the pixels in a symmetry map, wherein the symmetry map stores the symmetry value for each pixel in the image as a feature of the respective pixel, and wherein each symmetry value is dependent on the respective difference.

2. The method as recited in claim 1, wherein automatically determining the symmetry value for each of the plurality of pixels in the image comprises:

determining one or more interest points in the image;

determining a matching symmetric interest point for each interest point, wherein each interest point and matching symmetric interest point comprise a matching interest point pair; and

voting for the axis of symmetry for each matching interest point pair.

3. The method as recited in claim 1, wherein automatically determining the symmetry value for each of the plurality of pixels in the image further comprises:

determining a photometric distortion between the area around each of the plurality of pixels and the area around the corresponding symmetric pixel;

removing the photometric distortion for each of the plurality of pixels;

normalizing the difference between the area around each of the plurality of pixels and the area around the corresponding symmetric pixel.

4. The method as recited in claim 1, further comprising:

segmenting the image based on the symmetry values.

5. The method as recited in claim 1, further comprising:

annotating the image based on the symmetry values.

6. The method as recited in claim 1, further comprising:

determining one or more image primitives in the image based on the symmetry values.

7. The method as recited in claim 1, wherein the one or more symmetries comprise a bilateral symmetry in at least a portion of the image.

8. The method as recited in claim 1, wherein the one or more symmetries comprise a radial symmetry in at least a portion of the image.

9. The method as recited in claim 1, wherein the one or more symmetries comprise a periodic pattern in at least a portion of the image.

10. The method as recited in claim 1, wherein automatically determining the symmetry value for each of the plurality of pixels in the image comprises:

determining one or more axes of symmetry in the image;

determining one or more axes of symmetry in a down-sampled version of the image;

determining one or more axes of symmetry in an up-sampled version of the image; and

combining the one or more axes of symmetry in the image, the one or more axes of symmetry in the down-sampled version of the image, and the one or more axes of symmetry in the up-sampled version of the image.

11. The method as recited in claim 1, wherein automatically determining the symmetry value for each of the plurality of pixels in the image comprises automatically determining the symmetry value for each of the plurality of pixels in one or more down-sampled versions of the image or one or more up-sampled versions of the image.

12. A non-transitory, computer-readable storage medium, comprising program instructions, wherein the program instructions are computer-executable to implement:

determining a symmetry value for each of a plurality of pixels in an image, wherein the symmetry value indicates a strength of one or more symmetries in the image for the respective pixel with respect to one or more other pixels in the image, and wherein determining the symmetry value for each of the plurality of pixels in the image comprises;

determining a symmetric pixel for an axis of symmetry for each of the plurality of pixels; and

determining a respective difference between an area around each of the plurality of pixels and an area around the corresponding symmetric pixel, wherein the area around each of the plurality of pixels includes at least the respective pixel, and wherein the area around the corresponding symmetric pixel includes at least the corresponding symmetric pixel; and

storing the symmetry value for each of the pixels in a symmetry map, wherein the symmetry map stores the symmetry value for each pixel in the image as a feature of the respective pixel, and wherein each symmetry value is dependent on the respective difference.

13. The non-transitory, computer-readable storage medium as recited in claim 12, wherein automatically determining the symmetry value for each of the plurality of pixels in the image comprises:

determining one or more interest points in the image;

determining a matching symmetric interest point for each interest point, wherein each interest point and matching symmetric interest point comprise a matching interest point pair; and

voting for the axis of symmetry for each matching interest point pair.

14. The non-transitory, computer-readable storage medium as recited in claim 12, wherein automatically determining the symmetry value for each of the plurality of pixels in the image further comprises:

determining a photometric distortion between the area around each of the plurality of pixels and the area around the corresponding symmetric pixel;

removing the photometric distortion for each of the plurality of pixels;

normalizing the difference between the area around each of the plurality of pixels and the area around the corresponding symmetric pixel.

15. The non-transitory, computer-readable storage medium as recited in claim 12, wherein the program instructions are further computer-executable to implement:

segmenting the image based on the symmetry values.

16. The non-transitory, computer-readable storage medium as recited in claim 12, wherein the program instructions are further computer-executable to implement:

annotating the image based on the symmetry values.

17. The non-transitory, computer-readable storage medium as recited in claim 12, wherein the program instructions are further computer-executable to implement:

determining one or more image primitives in the image based on the symmetry values.

18. The non-transitory, computer-readable storage medium as recited in claim 12, wherein automatically determining the symmetry value for each of the plurality of pixels in the image comprises:

determining one or more axes of symmetry in the image;

determining one or more axes of symmetry in a down-sampled version of the image;

determining one or more axes of symmetry in an up-sampled version of the image; and

combining the one or more axes of symmetry in the image, the one or more axes of symmetry in the down-sampled version of the image, and the one or more axes of symmetry in the up-sampled version of the image.

19. The non-transitory, computer-readable storage medium as recited in claim 12, wherein automatically determining the symmetry value for each of the plurality of pixels in the image comprises automatically determining the symmetry value for each of the plurality of pixels in one or more down-sampled versions of the image or one or more up-sampled versions of the image.

20. A system, comprising:

at least one processor;

a memory coupled to the at least one processor, wherein the memory is configured to store program instructions executable by the at least one processor to;

determine a symmetry value for each of a plurality of pixels in an image, wherein the symmetry value indicates a strength of one or more symmetries in the image for the respective pixel with respect to one or more other pixels in the image, and wherein, in determining the symmetry value for each of the plurality of pixels in the image, the program instructions are executable by the at least one processor to;

determine a symmetric pixel for an axis of symmetry for each of the plurality of pixels; and

determine a respective difference between an area around each of the plurality of pixels and an area around the corresponding symmetric pixel, wherein the area around each of the plurality of pixels includes at least the respective pixel, and wherein the area around the corresponding symmetric pixel includes at least the corresponding symmetric pixel; and

store the symmetry value for each of the pixels in a symmetry map, wherein the symmetry map stores the symmetry value for each pixel in the image as a feature of the respective pixel, and wherein each symmetry value is dependent on the respective difference.

21. The system as recited in claim 20, wherein, in automatically determining the symmetry value for each of the plurality of pixels in the image, the program instructions are further executable by the at least one processor to:

determine one or more interest points in the image;

determine a matching symmetric interest point for each interest point, wherein each interest point and matching symmetric interest point comprise a matching interest point pair; and

vote for the axis of symmetry for each matching interest point pair.

22. The system as recited in claim 20, wherein, in automatically determining the symmetry value for each of the plurality of pixels in the image, the program instructions are further executable by the at least one processor to:

determine a photometric distortion between the area around each of the plurality of pixels and the area around the corresponding symmetric pixel;

remove the photometric distortion for each of the plurality of pixels;

normalize the difference between the area around each of the plurality of pixels and the area around the corresponding symmetric pixel.

23. The system as recited in claim 20, wherein the program instructions are further executable by the at least one processor to:

segment the image based on the symmetry values.

24. The system as recited in claim 20, wherein the program instructions are further executable by the at least one processor to:

annotate the image based on the symmetry values.

25. The system as recited in claim 20, wherein the program instructions are further executable by the at least one processor to:

determine one or more image primitives in the image based on the symmetry values.

26. The system as recited in claim 20, wherein, in automatically determining the symmetry value for each of the plurality of pixels in the image, the program instructions are further executable by the at least one processor to:

determine one or more axes of symmetry in the image;

determine one or more axes of symmetry in a down-sampled version of the image;

determine one or more axes of symmetry in an up-sampled version of the image; and

combine the one or more axes of symmetry in the image, the one or more axes of symmetry in the down-sampled version of the image, and the one or more axes of symmetry in the up-sampled version of the image.

27. The system as recited in claim 20, wherein, in automatically determining the symmetry value for each of the plurality of pixels in the image, the program instructions are further executable by the at least one processor to:

automatically determine the symmetry value for each of the plurality of pixels in one or more down-sampled versions of the image or one or more up-sampled versions of the image.