Real-time face tracking in a digital image acquisition device

• US 7,460,695 B2
• Filed: 06/21/2007
• Issued: 12/02/2008
• Est. Priority Date: 08/11/2006
• Status: Expired due to Fees

• Current Assignee: FotoNation Limited (Adeia Inc.)
• Original Assignee: FotoNation Vision Limited (Adeia Inc.)
• Inventors: Steinberg, Eran, Petrescu, Stefan, Nanu, Florin, Drimbarean, Alexandru, Bigioi, Petronel, Corcoran, Peter, Pososin, Alexei
• Primary Examiner(s): Alavi; Amir
• Application Number:
  US11/766,674
• Publication Number:
  US 20080037840A1
• Time in Patent Office: 530 Days
• Field of Search: 382/103, 382/115, 382/118, 382/164, 382/173, 382/190, 382/254, 382/276, 382/284, 382/298, 382/299, 382/305
• US Class Current: 382/118
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06F 18/2148 characterised by the process organisation or structure, e.g. boosting cascade
• CPC Class Codes: G06V 30/19147 Obtaining sets of training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06V 30/2504 Coarse or fine approaches, e.g. resolution of ambiguities or multiscale approaches
• CPC Class Codes: G06V 40/162 using pixel segmentation or colour matching
• CPC Class Codes: G06V 40/165 using facial parts and geometric relationships
• CPC Class Codes: G06V 40/167 using comparisons between temporally consecutive images
• CPC Class Codes: G06V 40/171 Local features and components; Facial parts eye characteristics G06V40/18; Occluding parts, e.g. glasses; Geometrical relationships
• CPC Class Codes: H04N 1/2112 using still video cameras
• CPC Class Codes: H04N 1/32112 in a separate computer file, document page or paper sheet, e.g. a fax cover sheet H04N1/32122 takes precedence
• CPC Class Codes: H04N 1/32128 attached to the image data, e.g. file header, transmitted message header, information on the same page or in the same computer file as the image for information embedded in the image H04N1/32144
• CPC Class Codes: H04N 2101/00 Still video cameras
• CPC Class Codes: H04N 2201/0084 Digital still camera
• CPC Class Codes: H04N 2201/3205 of identification information, e.g. name or ID code H04N2201/3209 takes precedence
• CPC Class Codes: H04N 23/61 based on recognised objects
• CPC Class Codes: H04N 23/611 where the recognised objects include parts of the human body
• CPC Class Codes: H04N 23/635 Region indicators; Field of view indicators
• CPC Class Codes: H04N 23/695 Control of camera direction for changing a field of view, e.g. pan, tilt or based on tracking of objects

Abstract

An image processing apparatus for tracking faces in an image stream iteratively receives an acquired image from the image stream potentially including one or more face regions. The acquired image is sub-sampled at a specified resolution to provide a sub-sampled image. An integral image is then calculated for a least a portion of the sub-sampled image. Fixed size face detection is applied to at least a portion of the integral image to provide a set of candidate face regions. Responsive to the set of candidate face regions produced and any previously detected candidate face regions, the resolution is adjusted for sub-sampling a subsequent acquired image.

Claims

1. A method of tracking faces in an image stream with a digital image acquisition device, comprising:

a. receiving an acquired image from an image stream including one or more face regions;

b. sub-sampling said acquired image at a specified resolution to provide a sub-sampled image;

c. calculating a corresponding integral image for a least a portion of said sub-sampled image;

d. applying face detection to at least a portion of said integral image to provide a set of one or more candidate face regions;

e. adjusting a resolution for sub-sampling a next acquired image;

f. repeating a. to e. at the adjusted resolution periodically on a selected plurality of images of an image stream including a main acquired image chronologically following a plurality of images of said selected plurality of images.

2. A method as claimed in claim 1, further comprising tracking candidate face regions of different sizes from the plurality of images of said image stream.

3. A method as claimed in claim 1, further comprising merging said set of candidate face regions with a previously detected face region to provide a set of candidate face regions of different sizes.

4. A method as claimed in claim 3, further comprising, for each region of said acquired image corresponding to a region of said merged set of candidate face regions:

calculating an integral image; and

applying variable-sized face detection to each merged region integral image to provide a set of confirmed face regions and a set of rejected face regions.

5. A method as claimed in claim 4, further comprising:

checking a rejected face region based on alternative criteria from said variable sized face detection; and

responsive to said checking indicating a rejected face region is a face region, adding said previously rejected face region to said set of confirmed face regions.

6. A method as claimed in claim 5, wherein said checking comprises applying a skin prototype to the rejected face region.

7. A method as claimed in claim 4, wherein said merging comprises merging a set of candidate face regions for an acquired image with said set of confirmed face regions for a previously acquired image.

8. A method as claimed in claim 4, wherein said regions of said acquired image corresponding to a region of said merged set of candidate face regions comprise regions surrounding respective regions of said merged set of candidate face regions.

9. A method as claimed in claim 4, wherein applying variable sized face detection comprises applying a plurality of cascades of Haar classifiers of varying size to each merged region integral image.

10. A method as claimed in claim 9, comprising applying twelve or fewer sizes of cascades of Haar classifiers.

11. The method of claim 9, comprising applying 3 or 4 sizes of Haar classifiers.

12. A method as claimed in claim 11, wherein each cacade comprises 32 classifiers.

13. A method as claimed in claim 1, wherein said adjusting comprises cycling through a set of at least approximately four sub-sampling resolutions.

14. A method as claimed in claim 1, wherein said adjusting is responsive to said image being acquired at approximately infinite focus for adjusting said sub-sampling resolution to maximum resolution.

15. A method as claimed in claim 1, further comprising storing size or location, or both, of a candidate face region in association with said main acquired image.

16. A method as claimed in claim 15, further comprising performing spatially selective post processing of said main acquired image based on said stored candidate face region'"'"'s size or location or both.

17. A method as claimed in claim 16, wherein said selective image processing comprises color correction, sharpening, blurring, saturation, subsampling, or compression, or a combination thereof.

18. A method as claimed in claim 1, wherein said face detection comprises a fixed size, and wherein said one or more candidate face regions each have a given size and respective location.

19. A method as claimed in claim 1, where the time interval between said selected images is fixed.

20. A method as claimed in claim 1, where the time interval between said selected images is variable and determined during execution.

21. A method as claim in claim 20, wherein said time interval is a function of the time taken to process one or more previous main acquired images.

22. A method as claimed in claim 21, wherein said time interval is dependent on changes in image content for one or more of said selected plurality of images.

23. A method as claimed in claim 21, wherein said cascade comprises at least approximately 32 classifiers.

24. A method as claimed in claim 1, wherein said repeating steps a. to e. is performed responsive to changes in content from image to image in said image stream.

25. A method as claimed in claim 1, wherein said applying face detection comprises applying a cascade of Haar classifiers of a fixed size to said integral image.

26. A method as claimed in claim 1, further comprising:

displaying an acquired image; and

superimposing said tracked candidate face regions on said displayed acquired image.

27. A method as claimed in claim 1, further comprising responsive to said acquired image being captured with a flash, analyzing regions of said acquired image corresponding to said tracked candidate face regions for red-eye defects.

28. A method as claimed in claim 27, further comprising correcting in said acquired image any red-eye defects.

29. A method as claimed in claim 27, further comprising storing with said acquired image an indication of any red-eye defects.

30. An image processing apparatus for tracking faces in an image stream comprising one or more processors and one or more digital storage media having digitally-encoded instructions embedded therein for programming the one or more processors to perform an iterative method of tracking faces in an image stream, the method comprising:

a. receiving an acquired image from an image stream including one or more face regions;

b. sub-sampling said acquired image at a specified resolution to provide a sub-sampled image;

c. calculating a corresponding integral image for a least a portion of said sub-sampled image;

d. applying face detection to at least a portion of said integral image to provide a set of one or more candidate face regions;

e. adjusting a resolution for sub-sampling a next acquired image;

f. repeating a. to e. at the adjusted resolution periodically on a selected plurality of images of an image stream including a main acquired image chronologically following a plurality of images of said selected plurality of images.

31. An apparatus according to claim 30, comprising a digital still camera, a video camera, a cell phone equipped with an image capturing mechanism or a hand help computer equipped with an internal or external camera, or combinations thereof.

32. A method of detecting faces in an image stream using a digital image acquisition device, comprising:

a. determining an orientation of a digital image acquisition device for at least one image of an image stream;

b. acquiring the image, which includes one or more face regions, from said image stream; and

c. applying face detection to at least a portion of said acquired image to provide a set of one or more candidate face regions according to said determined orientation, and wherein said face detection selectively employs a sequence of classifiers from an available plurality of classifier sequences, each classifier sequence being at a given orientation.

33. A method according to claim 32, wherein each candidate face region has a given size or respective location.

34. A method according to claim 32, further comprising:

selectively applying face recognition with said database to at least one candidate face region;

providing an identifier for each face recognized in a candidate face region; and

storing the identifier for each recognized face in association with at least one image of said image stream.

35. An image processing apparatus including one or more processors and one or more digital storage media having digitally-encoded instructions embedded therein for programming the one or more processors to perform an iterative method of detecting faces in an image stream, the method comprising:

a. determining an orientation of a digital image acquisition device for at least one image of an image stream;

b. acquiring the image, which includes one or more face regions, from said image stream; and

c. applying face detection to at least a portion of said acquired image to provide a set of one or more candidate face regions according to said determined orientation, and wherein said face detection selectively employs a sequence of classifiers from an available plurality of classifier sequences, each classifier sequence being at a given orientation.

36. An apparatus according to claim 35, wherein each candidate face region has a given size or respective location.

37. An apparatus according to claim 35, further comprising:

selectively applying face recognition with said database to at least one candidate face region;

providing an identifier for each face recognized in a candidate face region; and

storing the identifier for each recognized face in association with at least one image of said image stream.

38. An image processing apparatus according to claim 35, further comprising a hardware orientation sensor for indicating the orientation of said apparatus.

39. An image processing apparatus according to claim 35, wherein the method further comprises analyzing content in one or more images of said image stream to determine the orientation of said apparatus.

40. An image processing apparatus according to claim 35, further comprising:

selectively applying face recognition with said database to at least one candidate face region;

providing an identifier for each face recognized in a candidate face region; and

storing the identifier for each recognized face in association with at least one image of said image stream.

41. A method of detecting faces in an image stream using a digital image acquisition device comprising:

a. receiving a first acquired image from said image stream including one or more face regions;

b. applying face detection to at least a portion of said first acquired image to provide a set of one or more candidate face regions each having a probability of being a face region;

c. tracking candidate face regions across a plurality of images of said image stream including said first acquired image; and

d. responsive to a probability of a tracked face region dropping by more than a threshold rate with respect to time, activating an enhancement sub-system for a period of time to improve the probability of said face detection detecting a candidate face region.

42. A method according to claim 41 wherein said activating comprises:

acquiring each new image of said image stream with a progressively increased exposure parameter until at least one candidate face region is detected with a probability above a given threshold.

43. A method according to claim 41 wherein said activating comprises:

activating one or more of an LED, an infra-red LED, a visible light source, or a micro-flash; and

subsequently acquiring one or more images of said image stream until at least one candidate face region is detected with a probability above a given threshold.

44. A method according to claim 41 wherein said activating comprises:

changing an optical filter; and

subsequently acquiring one or more images of said image stream until at least one candidate face region is detected with a probability above a given threshold.

45. A method according to claim 41 wherein said activating comprises:

activating a second image acquisition sensor; and

subsequently acquiring one or more images of said image stream until at least one candidate face region is detected with a probability above a given threshold.

46. A method according to claim 41 wherein said applying face detection comprises:

sub-sampling said acquired image at a specified resolution to provide a sub-sampled image;

identifying one or more regions of said acquired image predominantly including skin tones; and

calculating a corresponding integral image for a least one of said skin tone regions of said sub-sampled acquired image; and

applying face detection to at least a portion of said integral image.

47. A method as claimed in claim 46, wherein said identifying is performed on said sub-sampled image.

48. A method as claimed in claim 41, further comprising:

prior to applying face detection for a candidate face region associated with a previous image, enhancing the contrast of the luminance characteristics of the corresponding region of said acquired image.

49. A method as claimed in claim 41 further comprising storing a history of a tracked region'"'"'s probability with an acquired image.

50. An apparatus as claimed in claim 41, wherein the method further comprises:

prior to applying face detection for a candidate face region associated with a previous image, enhancing the contrast of the luminance characteristics of the corresponding region of said acquired image.

51. A digital image acquisition device for detecting faces in an image stream comprising:

a. means for receiving a first acquired image from said image stream including one or more face regions;

b. means for applying face detection to at least a portion of said first acquired image to provide a set of one or more candidate face regions each having a probability of being a face region;

c. means for tracking candidate face regions across a plurality of images of said image stream including said first acquired image; and

d. means, responsive to a probability of a tracked face region dropping by more than a threshold rate with respect to time, for activating an enhancement sub-system for a period of time to improve the probability of said face detection detecting a candidate face region.

52. An image processing apparatus including one or more processors and one or more digital storage media having digitally-encoded instructions embedded therein for programming the one or more processors to perform a method of detecting faces in an image stream using a digital image acquisition device, the method comprising:

a. receiving a first acquired image from said image stream including one or more face regions;

b. applying face detection to at least a portion of said first acquired image to provide a set of one or more candidate face regions each having a probability of being a face region;

c. tracking candidate face regions across a plurality of images of said image stream including said first acquired image; and

d. responsive to a probability of a tracked face region dropping by more than a threshold rate with respect to time, activating an enhancement sub-system for a period of time to improve the probability of said face detection detecting a candidate face region.

53. An apparatus according to claim 52 wherein said activating comprises:

acquiring each new image of said image stream with a progressively increased exposure parameter until at least one candidate face region is detected with a probability above a given threshold.

54. An apparatus according to claim 52 wherein said activating comprises:

activating one or more of an LED, an infra-red LED, a visible light source, or a micro-flash; and

subsequently acquiring one or more images of said image stream until at least one candidate face region is detected with a probability above a given threshold.

55. An apparatus according to claim 52 wherein said activating comprises:

changing an optical filter; and

subsequently acquiring one or more images of said image stream until at least one candidate face region is detected with a probability above a given threshold.

56. An apparatus according to claim 52 wherein said activating comprises:

activating a second image acquisition sensor; and

subsequently acquiring one or more images of said image stream until at least one candidate face region is detected with a probability above a given threshold.

57. An apparatus as claimed in claim 52 wherein said applying face detection comprises:

sub-sampling said acquired image at a specified resolution to provide a sub-sampled image;

identifying one or more regions of said acquired image predominantly including skin tones; and

calculating a corresponding integral image for a least one of said skin tone regions of said sub-sampled acquired image; and

applying face detection to at least a portion of said integral image.

58. An apparatus as claimed in claim 57, wherein said identifying is performed on said sub-sampled image.

59. An apparatus as claimed in claim 52, wherein the method further comprises storing a history of a tracked region'"'"'s probability with an acquired image.