Image processing methods and apparatus for detecting human eyes, human face, and other objects in an image

• US 6,965,684 B2
• Filed: 09/14/2001
• Issued: 11/15/2005
• Est. Priority Date: 09/15/2000
• Status: Expired due to Fees

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Ji, Xin, Chen, Xinwu
• Primary Examiner(s): Johns, Andrew W.
• Application Number:
  US09/951,458
• Publication Number:
  US 20020081032A1
• Time in Patent Office: 1,523 Days
• Field of Search: 382/100, 382/103, 382/117, 382/173, 382/180, 382/270, 382/273, 348/78
• US Class Current: 382/103
• CPC Class Codes: G06T 2207/20132 Image cropping
• CPC Class Codes: G06T 2207/30201 Face
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06V 40/161 Detection; Localisation; Normalisation

Abstract

Disclosed is a method, apparatus, and system for detecting a human face in an image. The method includes, for a subset of pixels in the image, deriving a first variable from the gray-level distribution of the image and deriving a second variable from a preset reference distribution that is characteristic of the object. The method further includes evaluating the correspondence between the first variable and the second variable over the subset of pixels and determining if the image contains the object based on the result of this evaluation.

Claims

1. A human eye detecting method for detecting human eye in an image, comprising:

a reading step of reading the gray level of each pixel in the each column in the image;

a segmenting step of segmenting each column into a plurality of intervals, and labeling each of the intervals as valley region, intermediate region or peak region;

a merging step of merging the valley region of the each column and the valley region of its adjacent column, and generating an eye candidate region; and

a determining step of determining the human eye from the eye candidate regions.

2. The method according to claim 1, wherein the segmenting step includes the step of labeling each interval as one of valley region, intermediate region and peak region based on the gray level of each the interval in a column.

3. The method according to claim 1, wherein the segmenting step includes the step of labeling each interval as one of valley region, intermediate region and peak region based on the luminance value of each interval in a column.

4. The method according to claim 1, wherein the segmenting step includes the step of labeling each interval in a column as one of valley region, intermediate region and peak region based on the respective gray level ratios of the interval to its two adjacent intervals.

5. The method according to claim 1, wherein the segmenting step includes step of labeling each interval in a column as one of valley region, intermediate region and peak region based on the respective luminance value ratios of the interval to its two adjacent interval.

6. The method according to any one of claim 1, further comprising the step of comparing the gray level of a segment point with that of its two adjacent intervals, and adding the segment point into the adjacent interval which has a gray level close to that of the segment point.

7. The method according to any one of claim 1, further comprising the step of comparing the gray level of a intermediate region with that of its adjacent valley region and peak valley, and adding the intermediate region into the valley region on the basis of the gray levels of the adjacent valley region, peak region and the intermediate region.

8. The method according to claim 1, wherein the merging step comprising steps of

setting each valley region in the first column of the image as a seed region respectively;

determining whether the valley region of next column of the image can be merged into the seed region;

merging the valley region that can be merged into the seed region;

setting the non-merged valley region as a seed region; and

determining the seed region which has no further merged valley region as an eye candidate region.

9. The method according to claim 8, further comprising the step of comparing the size of overlap of the valley region and the seed region'"'"'s predicted region.

10. The method according to claim 8, further comprising the step of comparing the gray level of the valley region and that of the seed region.

11. The method according to claim 8, further comprising the step of comparing the gray levels of the valley region, the seed region, the background gray levels of the valley region and the seed region.

12. The method according to claim 1, wherein the determining step comprising the step of determining whether the gray level of the eye candidate region is less than the first threshold, where the first threshold is within the range of 100 to 200.

13. The method according to claim 12, wherein the first threshold is 160.

14. The method according to claim 12, wherein the determining step further comprising the step of determining whether the background gray level of the eye candidate region is bigger than the second threshold, where the second threshold is within the range of 20 to 80.

15. The method according to claim 14, wherein the second threshold is 30.

16. The method according to claim 14, wherein the determining step further comprising the step of determining whether the difference of background gray level of the eye candidate region and its gray level is bigger than a third threshold, where the third threshold is within the range of 5 to 120.

17. The method according to claim 16, wherein the third threshold is 20.

18. The method according to claim 16, wherein the determining step further comprising the step of determining whether the ratio of the width to the height of an eye candidate region is bigger than a fourth threshold, where the fourth threshold is within the range of 1 to 5.

19. The method according to claim 18, wherein the fourth threshold is 3.33.

20. The method according to claim 18, wherein the determining step further comprising the step of determining whether the ratio of the size of an eye candidate region to that of its circum-rectangle is bigger than the fifth threshold, where the fifth threshold is within the range of 0.2 to 1.

21. The method according to claim 20, wherein the fifth threshold is 0.4.

22. Apparatus for detecting a human eye in an image, comprising:

reading means for reading the gray level of each pixel in the each column in the image;

segmenting means for segmenting each column into a plurality of intervals, and labeling each of the intervals as valley region, intermediate region or peak region;

merging means for merging the valley region of the each column and the valley region of its adjacent column, and generating an eye candidate region; and

determining means for determining if a human eye exists in the eye candidate regions.

23. The apparatus according to claim 22, wherein the segmenting means includes means for labeling each interval as one of valley region, intermediate region and peak region based on, the gray level of each interval in a column.

24. The apparatus according to claim 22, wherein the segmenting means includes means for labeling each interval as one of valley region, intermediate region and peak region based on the luminance value of each interval in a column.

25. The apparatus according to claim 22, wherein the segmenting means includes means for labeling each interval in a column as one of valley region, intermediate region and peak region based on the respective gray level ratios of the interval to its two adjacent intervals.

26. The apparatus according to claim 22, wherein the segmenting means includes means for labeling each interval in a column as one of valley region, intermediate region and peak region based on the respective luminance value ratios of the interval to its two adjacent intervals.

27. The apparatus according to claim 22, further comprising means for comparing the gray level of a segment point with that of its two adjacent intervals, and adding the segment point into the adjacent interval which has a gray level close to that of the segment point.

28. The apparatus according to claim 22, further comprising means for comparing the gray level of an intermediate region with that of its adjacent valley region and peak valley, and adding the intermediate region into the valley region on the basis of the gray levels of the adjacent valley region, peak region and the intermediate region.

29. The apparatus according to claim 22, wherein the merging means comprises:

means for setting each valley region in the first column of the image as a seed region respectively;

means for determining whether the valley region of a next column of the image can be merged into the seed region;

means for merging a valley region that can be merged into the seed region;

means for setting a non-merged valley region as a seed region; and

means for determining the seed region which has no further merged valley region as an eye candidate region.

30. The apparatus according to claim 29, further comprising means for comparing the size of overlap of the valley region and the seed region'"'"'s predicted valley region.

31. The apparatus according to claim 29, further comprising means for comparing the gray level of the valley region and that of the seed region.

32. The apparatus according to claim 29, further comprising means for comparing the gray levels of the valley region, the seed region, the background gray levels of the valley region and the seed region.

33. The apparatus according to claim 22, wherein the determining means comprises means for determining whether the gray level of the eye candidate region is less than a first threshold, where the first threshold is within the range of 100 to 200.

34. The apparatus according to claim 33, wherein the first threshold is 160.

35. The apparatus according to claim 33 or claim 34, wherein the determining means further comprises means for determining whether the background gray level of the eye candidate region is bigger than a second threshold, where the second threshold is within the range 20 to 80.

36. The apparatus according to claim 35, wherein the second threshold is 30.

37. The apparatus according to claim 35 or claim 36, wherein the determining means further comprises means for determining whether the difference of background gray level of the eye candidate region and its gray level is bigger than a third threshold, where the third threshold is within the range 5 to 120.

38. The apparatus according to claim 37, wherein the third threshold is 20.

39. The apparatus according to claim 37 or claim 38, wherein the determining means further comprises means for determining whether the ratio of the width to the height of an eye candidate region is bigger than a fourth threshold, where the fourth threshold is within the range of 1 to 5.

40. The apparatus according to claim 39, wherein the fourth threshold is 3.33.

41. The apparatus according to claim 39 or claim 40, wherein the determining means further comprises means for determining whether the ratio of the size of an eye candidate region to that of its circum-rectangle is bigger than a fifth threshold, where the fifth threshold is within the range of 0.2 to 1.

42. The apparatus according to claim 41, wherein the fifth threshold is 0.4.