Method and system for determining object pose from images

• US 20060269145A1
• Filed: 04/08/2004
• Published: 11/30/2006
• Est. Priority Date: 04/17/2003
• Status: Abandoned Application

• Current Assignee: University of Dundee
• Original Assignee: University of Dundee
• Inventors: Roberts, Timothy
• Application Number:
  US10/553,664
• Publication Number:
  US 20060269145A1
• Time in Patent Office: Days
• US Class Current: 382/228
• CPC Class Codes: G06V 10/422 for representing the structure of the pattern or shape of an object therefor
• CPC Class Codes: G06V 40/103 Static body considered as a whole, e.g. static pedestrian or occupant recognition

Abstract

A method and system for identifying an object or structured parts of an object in an image. A set of templates are created for each of a number of the parts of the object and the templates are applied to an area of interest in an image where it is hypothesised that an object part is present. The image is analysed to determine the probability that it contains the object part. Thereafter, other templates are applied to other areas of interest in the image to determine the probability that this area of interest belongs to a corresponding object part. The templates are then arranged in a configuration and the likelihood that the configuration represents an object or structured parts of an object is calculated. This is calculated for other configurations and the configuration that is most likely to represent an object or structured part of an object is determined. The method and system can be applied to creating a markerless motion capture system and has other applications in image processing.

Claims

1. A method of identifying an object or structured parts of an object in an image, the method comprising the steps of:

creating a set of templates, the set containing a template for each of a number of predetermined object parts and applying said template to an area of interest in an image where it is hypothesised that an object part is present;

analysing image pixels in the area of interest to determine the probability that it contains the object part;

applying other templates from the set of templates to other areas of interest in the image to determine the probability that said area of interest belongs to a corresponding object part and arranging the templates in a configuration;

calculating the likelihood that the configuration represents an object or structured parts of an object; and

calculating other configurations and comparing said configurations to determine the configuration that is most likely to represent an object or structured part of an object.

2. A method as claimed in claim 1 wherein, the probability that an area of interest contains an object part is calculated by calculating a transformation from the co-ordinates of a pixel in the area of interest to the template.

3. A method as claimed in claim 1 wherein, analysing the area of interest further comprises identifying the dissimilarity between foreground and background of a transformed probabilistic region.

4. A method as claimed in claim 1 wherein, analysing the area of interest further comprises calculating a likelihood ratio based on a determination of the dissimilarity between foreground and background features of a transformed template.

5. A method as claimed in claim 1 wherein, the templates are applied by aligning their centres, orientations in 2D or 3D and scales to the area of interest on the image.

6. A method as claimed in claim 1 wherein the template is a probabilistic region mask in which values indicate a probability of finding a pixel corresponding to an object part.

7. A method as claimed in claim 1 wherein, the probabilistic region mask is estimated by segmentation of training images.

8. A method as claimed in claim 1 wherein, the image is an unconstrained scene.

9. A method as claimed in claim 1 wherein, the step of calculating the likelihood that the configuration represents an object or a structured part of an object comprises calculating a likelihood ratio for each object part and calculating the product of said likelihood ratios.

10. A method as claimed in claim 1 wherein, the step of calculating the likelihood that the configuration represents an object comprises determining the spatial relationship of object part templates.

11. A method as claimed in claim 10 wherein the step of determining the spatial relationship of the object part templates comprises analysing the configuration to identify common boundaries between pairs of object part templates.

12. A method as claimed in claim 11 wherein the step of determining the spatial relationship of the object part templates requires identification of object parts having similar characteristics and defining these as a sub-set of the object part templates.

13. A method as claimed in claim 12, wherein the step of calculating the likelihood that the configuration represents an object or structured part of an object comprises calculating a link value for object parts which are physically connected.

14. A method as claimed in claim 1 wherein the step of comparing said configurations comprises iteratively combining the object parts and predicting larger configurations of body parts.

15. A method as claimed in claim 1 wherein the object is a human or animal body.

16. A system for identifying an object or structured parts of an object in an image, the system comprising:

a set of templates, the set containing a template for each of a number of predetermined object parts applicable to an area of interest in an image where it is hypothesised that an object part is present;

analysis means for determining the probability that the area of interest contains the object part;

configuring means capable of arranging the applied templates in a configuration;

calculating means to calculate the likelihood that the configuration represents an object or structured parts of an object for a plurality of configurations; and

comparison means to compare configurations so as to determine the configuration that is most likely to represent an object or structured part of an object.

17. A system as claimed in claim 16 wherein, the system further comprises imaging means capable of providing an image for analysis.

18. A system as claimed in claim 17 wherein the imaging means is a stills camera or a video camera.

19. A system as claimed in claim 18 wherein, the analysis means is provided with means for identifying the dissimilarity between foreground and background of a transformed probabilistic region.

20. A system as claimed in claim 19 wherein, the analysis means calculates the probability that an area of interest contains an object part by calculating a transformation from the co-ordinates of a pixel in the area of interest to the template.

21. A system as claimed in claim 16 wherein, the analysis means calculates a likelihood ratio based on a determination of the dissimilarity between foreground and background features of a transformed template.

22. A system as claimed in claim 16 wherein, the templates are applied by aligning their centres, orientations (in 2D or 3D) and scales to the area of interest on the image.

23. A system as claimed in claim 16 wherein the template is a probabilistic region mask in which values indicate a probability of finding a pixel corresponding to the body part.

24. A system as claimed in claim 16 wherein, the probabilistic region mask is estimated by segmentation of training images.

25. A system as claimed in claim 16 wherein, the image is an unconstrained scene.

26. A system as claimed in claim 16 wherein, the calculating means calculates a likelihood ratio for each object part and calculating the product of said likelihood ratios.

27. A system as claimed in claim 26 wherein, the likelihood that the configuration represents an object comprises determining the spatial relationship of object part templates.

28. A system as claimed in claim 27 wherein the spatial relationship of the object part templates is calculated by analysing the configuration to identify common boundaries between pairs of object part templates.

29. A system as claimed in claim 28 wherein the spatial relationship of the object part templates is determined by identifying object parts having similar characteristics and defining these as a sub-set of the object part templates.

30. A system as claimed in claim 28, wherein the calculating means is capable of calculating a link value for object parts which are physically connected.

32. A system as claimed in claim 16, wherein the calculating means is capable of iteratively combining the object parts in order to predict larger configurations of body parts.

31. (canceled)

33. (canceled)

34. A computer program comprising program instructions for causing a computer to perform the method of

creating a set of templates the set containing a template for each of a number of predetermined object parts and applying said template to an area of interest in an image where it is hypothesised that an object part is present;

analysing image pixels in the area of interest to determine the probability that it contains the object part;

applying other templates from the set of templates to other areas of interest in the image to determine the probability that said area of interest belongs to a corresponding object part and arranging the templates in a configuration;

calculating the likelihood that the configuration represents an object or structured parts of an object; and

calculating other configurations and comparing said configurations to determine the configuration that is most likely to represent an object or structured part of an object.

35. A computer program as claimed in claim 34 wherein the computer program is embodied on a computer readable medium.

36. (canceled)

37. A markerless motion capture system comprising imaging means and a system for identifying an object or structured parts of an object in an image wherein the system includes:

a set of templates, the set containing a template for each of a number of predetermined object parts

applicable to an area of interest in an image where it is hypothesised that an object part is present;

analysis means for determining the probability that the area of interest contains the object part;

configuring means capable of arranging the applied templates in a configuration;

calculating means to calculate the likelihood that the configuration represents an object or structured parts of an object for a plurality of configurations; and

comparison means to compare configurations so as to determine the configuration that is most likely to represent an object or structured part of an object.