Image processing

• US 8,340,412 B2
• Filed: 10/27/2006
• Issued: 12/25/2012
• Est. Priority Date: 10/31/2005
• Status: Expired due to Fees

• Current Assignee: Sony United Kingdom Limited (Sony Group Corp.)
• Original Assignee: Sony United Kingdom Limited (Sony Group Corp.)
• Inventors: Lang, Mikael Carl, Porter, Robert Mark Stefan
• Primary Examiner(s): Sherali, Ishrat I
• Application Number:
  US12/092,206
• Publication Number:
  US 20100158362A1
• Time in Patent Office: 2,251 Days
• Field of Search: 382/162, 382/165, 382/108, 382/190, 382/305, 382224-228, 382/168, 382/170, 345/582, 345/589
• US Class Current: 382/165
• CPC Class Codes: G06F 16/5838 using colour
• CPC Class Codes: G06F 16/5854 using shape and object relationship
• CPC Class Codes: G06F 16/5862 using texture

Abstract

An image processing method includes detecting statistical distribution of permutations of pixel properties amongst contiguous groups of a predetermined number of pixels in at least a portion of an image under test, so as to derive feature data in respect of the image under test, and comparing the feature data derived in respect of the image under test with feature data derived from at least respective portions of one or more other images so as to detect a similarity between the image under test and the one or more other images.

Claims

1. An image processing method comprising:

quantizing pixel properties amongst contiguous groups of a predetermined number of pixels in at least a portion of an image under test to obtain quantized pixel property values;

detecting contributions of the pixel properties to the quantized pixel property values;

detecting, after the quantizing, a statistical distribution of permutations of the pixel properties and weighting the permutations based on the contributions to derive feature data from at least the portion of the image under test; and

comparing the feature data derived from at least the portion of the image under test with feature data derived from at least respective portions of one or more other images to detect a degree of similarity between the image under test and the one or more other images, wherein

the pixel properties are color properties.

2. The method according to claim 1, in which the detecting the statistical distribution of permutations includes detecting permutations of pixel properties amongst all possible contiguous groups of the predetermined number of pixels within at least the portion of the image under test.

3. The method according to claim 1, in which the predetermined number is two, so that the groups of pixels comprise pairs of adjacent pixels.

8. The method according to claim 1, further comprising:

grouping two or more images or image portions together if their feature data indicates greater than a predetermined degree of similarity.

16. The method according to claim 1, wherein the image under test is a medical image or a security image.

4. An image processing method comprising:

detecting a statistical distribution of permutations of pixel properties amongst contiguous groups of a predetermined number of pixels in at least a portion of an image under test;

generating occurrence frequency data of possible permutations of group pixel properties;

concatenating the occurrence frequency data to generate a feature vector as feature data derived from at least the portion of the image under test; and

comparing the feature data derived from at least the portion of the image under test with feature data derived from at least respective portions of one or more other images to detect a degree of similarity between the image under test and the one or more other images.

5. The method according to claim 4, further comprising:

reducing a size of the feature vector by multiplying the feature vector by a random number matrix.

6. The method according to claim 4, further comprising:

reducing a size of the feature vector by principal component analysis.

7. The method according to claim 4, wherein the comparing includes detecting a vector distance between the feature vector and a feature vector of the one or more other images.

17. The method according to claim 4, wherein the image under test is a medical image or a security image.

18. The method according to claim 4, further comprising:

grouping two or more images or image portions together if their feature data indicates greater than a predetermined degree of similarity.

19. The method according to claim 4, in which the detecting the statistical distribution of permutations includes detecting permutations of pixel properties amongst all possible contiguous groups of the predetermined number of pixels within at least the portion of the image under test.

9. A non-transitory computer readable medium comprising program code which, when executed by a computer, executes a method comprising:

quantizing pixel properties amongst contiguous groups of a predetermined number of pixels in at least a portion of an image under test to obtain quantized pixel property values;

detecting contributions of the pixel properties to the quantized pixel property values;

detecting, after the quantizing, a statistical distribution of permutations of the pixel properties and weighting the permutations based on the contributions to derive feature data from at least the portion of the image under test; and

comparing the feature data derived from at least the portion of the image under test with feature data derived from at least respective portions of one or more other images to detect a degree of similarity between the image under test and the one or more other images, wherein

the pixel properties are color properties.

10. An image processing apparatus comprising:

a quantizer to quantize pixel properties amongst contiguous groups of a predetermined number of pixels in at least a portion of an image under test to obtain quantized pixel property values;

a detector to detect contributions of the pixel properties to the quantized pixel property values, to detect a statistical distribution of permutations of the pixel properties, and to weight the permutations based on the contributions to derive feature data from at least the portion of the image under test; and

a comparator to compare the feature data derived from at least the portion of the image under test with feature data derived from at least respective portions of one or more other images to detect a degree of similarity between the image under test and the one or more other images, wherein

the pixel properties are color properties.

11. An image capture apparatus comprising the image processing apparatus according to claim 10.

12. The image processing apparatus according to claim 10, wherein the detector detects permutations of pixel properties amongst all possible contiguous groups of the predetermined number of pixels within at least the portion of the image under test.

13. The image processing apparatus according to claim 10, wherein the predetermined number is two, so that the groups of pixels comprise pairs of adjacent pixels.

14. The image processing apparatus according to claim 10, wherein the image under test is a medical image or a security image.

15. The image processing apparatus according to claim 10, further comprising:

a processor to group two or more images or image portions together if their feature data indicates greater than a predetermined degree of similarity.

20. An image processing apparatus comprising:

a detector to detect a statistical distribution of permutations of pixel properties amongst contiguous groups of a predetermined number of pixels in at least a portion of an image under test;

a generator to generate occurrence frequency data of possible permutations of group pixel properties;

a processor to concatenate the occurrence frequency data to generate a feature vector as feature data derived from at least the portion of the image under test; and

a comparator to compare the feature data derived from at least the portion of the image under test with feature data derived from at least respective portions of one or more other images to detect a degree of similarity between the image under test and the one or more other images.

21. An image capture apparatus comprising the image processing apparatus according to claim 20.

22. The image processing apparatus according to claim 20, wherein the detector detects permutations of pixel properties amongst all possible contiguous groups of the predetermined number of pixels within at least the portion of the image under test.

23. The image processing apparatus according to claim 20, further comprising:

a processor to reduce a size of the feature vector by multiplying the feature vector by a random number matrix.

24. The image processing apparatus according to claim 20, further comprising:

a processor to reduce a size of the feature vector by principal component analysis.

25. The image processing apparatus according to claim 20, wherein the comparator detects, in performing a comparison, a vector distance between the feature vector and a feature vector of the one or more other images.

26. The image processing apparatus according to claim 20, wherein the image under test is a medical image or a security image.

27. The image processing apparatus according to claim 20, further comprising:

a processor to group two or more images or image portions together if their feature data indicates greater than a predetermined degree of similarity.

28. A non-transitory computer readable medium comprising program code which, when executed by a computer, executes a method comprising:

detecting a statistical distribution of permutations of pixel properties amongst contiguous groups of a predetermined number of pixels in at least a portion of an image under test;

generating occurrence frequency data of possible permutations of group pixel properties;

concatenating the occurrence frequency data to generate a feature vector as feature data derived from at least the portion of the image under test; and

comparing the feature data derived from at least the portion of the image under test with feature data derived from at least respective portions of one or more other images to detect a degree of similarity between the image under test and the one or more other images.