Method of human figure contour outlining in images

1. A digital image processing method for automatically outlining a contour of the figure of a human body in a digital image, comprising the steps of:

a) testing parameters of a region within the digital image according to a plurality of cascaded tests;

b) determining whether the region contains characteristic features of the human body figure within the digital image;

c) computing location parameters of the characteristic features in the region for the figure of a human body within the digital image, where the region is related to the human body figure;

d) determining boundary parameters for the figure of a human body corresponding to the location parameters of the characteristic features in the region;

e) computing an information map of the digital image;

f) computing a set of indicative pixels for the contour of the figure of a human body; and

g) automatically outlining the contour of the human body figure using the set of indicative pixels, the information map, and a contour outlining tool.

2. The method claimed in claim 1 , wherein the step of computing a set of indicative pixels for the contour of the figure of a human body, includes the steps of:

f1) locating M search windows using relevant body measurements in a cost map for computing the set of indicative pixels for contour of the human body figure;

f2) collecting N cost map pixels with low cost values for each window;

f3) computing the location of a relevant human body figure contour point which is the statistical average location of the N pixels in each of the windows; and

f4) organizing computed relevant human body figure contour points, from each of the windows, as seed and target pixel pairs.

3. The method claimed in claim 2 , wherein the step of automatically outlining the contour of the human body figure using the set of indicative pixels and the contour outlining tool, includes the steps of:

g1) automatically generating pairs of pixels as seed and target end points, respectively, from the cost map, and;

g2) dynamically searching with the contour outlining tool to find an optimal path, relative to the cost map, from the seed to the target end points.

4. The method claimed in claim 1 , wherein the contour outlining tool is a modified non-parametric boundary finding algorithm.

5. The method claimed in claim 1 , wherein the step of computing an information map of the digital image further comprises employing intensity values of the digital image to generate a cost map.

6. The method claimed in claim 5 , wherein the intensity values of the digital image are used to find edges of the digital image for generating the cost map.

7. The method claimed in claim 4 , wherein the modified non-parametric boundary finding algorithm finds cost-minimized paths between a plurality of seed and target pixel pairs along the contour of the human body figure.

8. A digital image processing method for automatically outlining a contour of a human body figure in a digital image, comprising the steps of:

a) finding parameters of a facial region, corresponding to the human body figure within the digital image, according to a plurality of cascaded tests;

b) determining whether the facial region contains characteristic features of the human body figure within the digital image;

c) computing location parameters of the characteristic features in the facial region for the human body figure within the digital image;

d) determining boundary parameters for the human body figure corresponding to the location parameters of the characteristic features in the facial region;

e) computing an information map of the digital image;

f) computing a set of indicative pixels for the contour of the human body figure; and

g) automatically outlining the contour of the human body figure using the set of indicative pixels, the information map, and a contour outlining tool.

9. The method claimed in claim 8 , wherein the step of finding parameters of the facial region of the human body figure in the digital image employs a face detection algorithm.

10. The method claimed in claim 8 , wherein the step of computing location parameters of the characteristic features in the facial region employs an eye detection algorithm.

11. The method claimed in claim 8 , wherein the step of computing an information map of the digital image computes a cost map.

12. The method claimed in claim 10 , wherein the step of determining boundary parameters for the figure of a human body computes statistically relevant body measurements of the human body figure corresponding to locations of two eyes in the facial region.

13. The method claimed in claim 12 , wherein the step of computing a set of indicative pixels for contour of the human body figure, includes the steps of:

f1) locating M search windows using relevant body measurements in a cost map for computing the set of indicative pixels for contour of the human body figure;

f2) collecting N cost map pixels with low cost values for each window;

f3) computing a location of a relevant human body figure contour point which is the statistical average location of the N pixels in each of the windows; and

f4) organizing computed relevant human body figure contour points, from each of the windows, as seed and target pixel pairs.

14. The method claimed in claim 13 , wherein the step of automatically outlining the contour of the human body figure using the set of indicative pixels and the contour outlining tool, includes the steps of:

g1) automatically generating pairs of pixels as seed and target end points, respectively, from the cost map, and;

g2) dynamically searching with the contour outlining tool to find an optimal path, relative to the cost map, from the seed to the target end points.

15. The method claimed in claim 8 , wherein the contour outlining tool is a modified non-parametric boundary finding algorithm.

16. The method claimed in claim 8 , wherein the step of computing an information map of the digital image further comprises employing intensity values of the digital image to generate a cost map.

17. The method claimed in claim 16 , wherein the intensity values of the digital image are used to find edges of the digital image for generating the cost map.

18. The method claimed in claim 15 , wherein the modified non-parametric boundary finding algorithm finds cost-minimized paths between a plurality of seed and target pixel pairs along the contour of the human body figure.

19. The method claimed in claim 9 , wherein the face detection algorithm includes cascading a plurality of tests selected from the group consisting of: a chromaticity test, a geometry test, a grid pattern test, a location test, and a merging process.

20. The method claimed in claim 10 , wherein the eye detection algorithm includes the steps of:

a) clustering iris colored pixels as iris clusters;

b) validating the iris clusters;

c) counting the number of valid iris clusters as “n”, where “n”<2, employing template matching eye detection using image pixels;

d) where “n”≧2, employing template-less eye detection corresponding to the valid iris clusters;

e) determining whether eyes are detected; and

f) employing template matching eye detection corresponding to the valid iris clusters, when the template-less eye detection is unable to find a pair of eyes.

21. A digital image processing system for automatically outlining a contour of a human body figure in a digital image, comprising:

a) means for testing parameters of a region within the digital image according to a plurality of cascaded tests;

b) means for determining whether the region contains characteristic features of the human body figure within the digital image;

c) means for computing location parameters of the characteristic features in the region for the human body figure within the digital image;

d) means for determining boundary parameters for the human body figure corresponding to the location parameters of the characteristic features in the region;

e) means for computing an information map of the digital image;

f) means for computing a set of indicative pixels for the contour of the human body figure; and

g) means for automatically outlining the contour of the human body figure using the set of indicative pixels, the information map, and a contour outlining tool.

22. The system claimed in claim 21 , wherein the means for testing parameters of the region of the human body figure in the digital image employs a face detection algorithm.

23. The system claimed in claim 21 , wherein the means for computing location parameters of the characteristic features of the human body figure in the region within the digital image employs an eye detection algorithm.

24. The system claimed in claim 21 , wherein the means for computing an information map of the digital image computes a cost map.

25. The system claimed in claim 21 , wherein the means for determining boundary parameters for the human body figure computes statistically relevant body measurements of the human body figure corresponding to the location parameters of the characteristic features in the region.

26. A contoured digital image of a human body figure outlined according to the method of claim 1 .

27. A contoured digital image of a human body figure outlined according to the method of claim 8 .

28. A contour outlining tool for automatically outlining a contour of a human body figure within a digital image, in combination with a set of indicative pixels, comprising:

a) a detector that tests parameters of a region within the digital image according to a plurality of cascaded tests;

b) a first analyzer that determines whether the region contains characteristic features of the human body figure within the digital image;

c) a location processor that computes location parameters of the characteristic features in the region for the human body figure within the digital image;

d) a second analyzer that determines boundary parameters for the human body figure corresponding to the location parameters of the characteristic features in the region;

e) a mapping generator that generates an information map of the digital image relative to the boundary parameters of the human body figure; and

f) a pixel generator that generates the set of indicative pixels for the contour of the figure.

29. The contour outlining tool claimed in claim 28 , wherein the detector for testing parameters of the region of the human body figure in the digital image employs a face detection algorithm.

30. The contour outlining tool claimed in claim 28 , wherein the location processor for computing location parameters of the characteristic features of the human body figure in the region within the digital image employs an eye detection algorithm.

31. The contour outlining tool claimed in claim 28 , wherein the mapping generator for generating an information map of the digital image computes a cost map.

32. The contour outlining tool claimed in claim 28 , wherein the second analyzer for determining boundary parameters for the human body figure computes statistically relevant body measurements of the human body figure corresponding to the location parameters of the characteristic features in the region.

33. A contour outlining tool for automatically outlining a contour of a human body figure in a digital image, comprising:

a) a face detector that tests parameters of a facial region, corresponding to the human body figure within the digital image, according to a plurality of cascaded tests;

b) a first analyzer that determines whether the facial region contains characteristic features of the human body figure within the digital image;

c) a location processor that computes location parameters of the characteristic features in the facial region for the human body figure within the digital image;

d) a second analyzer that determines boundary parameters for the human body figure corresponding to the location parameters of the characteristic features in the facial region;

e) a map generator that generates an information map of the digital image relative to the boundary parameters of the human body figure; and

f) a pixel generator that generates a set of indicative pixels for the contour of the human body figure.

34. The contour outlining tool claimed in claim 33 , wherein the face detector employs a face detection algorithm.

35. The contour outlining tool claimed in claim 33 , wherein the location processor for computing location parameters of the characteristic features in the facial region employs an eye detection algorithm.

36. The contour outlining tool claimed in claim 33 , wherein the map generator computes a cost map.

37. The contour outlining tool claimed in claim 35 , wherein the second analyzer for determining boundary parameters for the human body figure computes statistically relevant body measurements of the human body figure corresponding to location of two eyes in the facial region.

38. The contour outlining tool claimed in claim 33 , further comprising a modified non-parametric boundary finding algorithm that finds cost-minimized paths between a plurality of seed and target pixel pairs along the contour of the human body figure.