Method and apparatus for processing an image

• US 20040131236A1
• Filed: 11/12/2003
• Published: 07/08/2004
• Est. Priority Date: 12/30/2002
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Ji, Xin, Wang, Libing, Chen, Xinwu
• Granted Patent: US 7,403,636 B2
• Time in Patent Office: Days
• US Class Current: 382/118
• CPC Class Codes: G06V 40/165 using facial parts and geometric relationships

Abstract

The present invention provides a method of processing an image, characterized by comprising steps of: identifying one candidate for human face region within said image; selecting a mouth neighborhood within said candidate for human face region; processing said mouth neighborhood; and classifying said candidate for human face region based on results of said processing step. According to the method of the present invention, human faces are detected just based on pixels included in the mouth neighborhood, but not all the pixels of the entire face.

Claims

1. A method of processing an image, characterized by comprising steps of:

identifying one candidate for human face region within said image;

selecting a mouth neighborhood within said candidate for human face region;

processing said mouth neighborhood; and

classifying said candidate for human face region based on results of said processing step.

2. The method of processing an image according to claim 1, characterized in that said step of selecting selects a mouth neighborhood which is r3 times the width of said candidate for human face region, and r1 times the height of said candidate for human face region, wherein r3 is a constant greater than 0.2 and less than 1, and r1 is a constant greater than 0.1 and less than 0.7.

3. The method of processing an image according to claim 2, characterized in that said r3 equals to 0.7, and said r1 equals to 0.3.

4. The method of processing an image according to any one of claims 1 to 3, characterized in that said step of processing comprises steps of:

forming an edge map at least corresponding to said mouth neighborhood;

calculating edge information for said mouth neighborhood based on said edge map;

and in that said step of classifying classifies said candidate for human face region based on said edge information.

5. The method of processing an image according to claim 4, characterized in that said step of calculating edge information comprises steps of:

selecting from a portion of said edge map which corresponds to said neighborhood pixels whose characteristic values are greater than a first threshold, resulting in a series of bright areas composed of pixels such selected;

calculating the size of said candidate for human face S1 and the size of the biggest bright area S2;

and in that said step of classifying comprises a step of;

classifying said candidate for human face region as a false human face, or one candidate with high possibility of being a false human face if S2/S1 is less than a second threshold.

6. The method of processing an image according to claim 5, characterized in that said step of calculating edge information further comprises a step of:

calculating the size of the second biggest bright area S3;

and in that said step of classifying further comprises a step of;

classifying said candidate for human face region as a false human face, or one candidate with high possibility of being a false human face if S3/S2 is less than a third threshold and the center of the biggest bright area is not in a mouth area, said mouth area being a predefined portion of said mouth neighborhood.

7. The method of processing an image according to claim 6, characterized in that said step of classifying further comprises a step of:

classifying said candidate for human face region as a real human face, or one candidate with high possibility of being a real human face if S3/S2 is less than said third threshold and said center of said biggest bright area is in said mouth area.

8. The method of processing an image according to claim 5, characterized in that said step of calculating edge information further comprises a step of:

calculating the size of the second biggest bright area S3;

and in that said step of classifying further comprises a step of;

classifying said candidate for human face region as a false human face, or one candidate with high possibility of being a false human face if S3/S2 is not less than a third threshold and the mid-point between the centers of the first two biggest bright areas is not in a mouth area, said mouth area being a predefined portion of said mouth neighborhood.

9. The method of processing an image according to claim 8, characterized in that said step of classifying further comprises a step of:

classifying said candidate for human face region as a real human face, or one candidate with high possibility of being a real human face if S3/S2 is not less than said third threshold and said mid-point between said centers of said first two biggest bright areas is in said mouth area.

10. The method of processing an image according to claim 4, characterized in that said step of calculating edge information comprises a step of:

calculating the average edge intensity within said mouth neighborhood and the average edge intensity within a mouth area, said mouth area being a predefined portion of said mouth neighborhood;

and in that said step of classifying comprises a step of;

classifying said candidate for human face region as a false human face, or one candidate with high possibility of being a false human face if the difference between said average edge intensity within said mouth area and said average edge intensity within said mouth neighborhood is not greater than a fourth threshold.

11. The method of processing an image according to claim 10, characterized in that said step of classifying further comprises a step of:

classifying said candidate for human face region as a real human face, or one candidate with high possibility of being a real human face if said difference between said average edge intensity within said mouth area and said average edge intensity within said mouth neighborhood is greater than said fourth threshold.

12. The method of processing an image according to any one of claims 6 to 9 and 10 to 11, characterized in that said mouth area is r2 times the width of said candidate for human face region, and at most r1 times the height of said candidate for human face region, wherein r2 is a constant greater than 0.1 and less than 0.7, and r1 is a constant greater than 0.1 and less than 0.7.

13. The method of processing an image according to claim 12, characterized in that said r2 equals to 0.5, and said r1 equals to 0.3.

14. An apparatus for processing an image, characterized by comprising:

a candidate identifier, for identifying one candidate for human face region within said image;

a mouth neighborhood selector, for selecting a mouth neighborhood within said candidate for human face region that has been identified by said candidate identifier;

a mouth neighborhood processor, for processing said mouth neighborhood that has been selected by said mouth neighborhood selector; and

a classifier, for classifying said candidate for human face region that has been identified by said candidate identifier based on outputs of said mouth neighborhood processor.

15. The apparatus for processing an image according to claim 14, characterized in that said mouth neighborhood selector selects a mouth neighborhood which is r3 times the width of said candidate for human face region, and r1 times the height of said candidate for human face region, wherein r3 is a constant greater than 0.2 and less than 1, and r1 is a constant greater than 0.1 and less than 0.7.

16. The apparatus for processing an image according to claim 15, characterized in that said r3 equals to 0.7, and said r1 equals to 0.3.

17. The apparatus for processing an image according to any one of claims 14 to 16, characterized in that said mouth neighborhood processor comprises:

a converter, for converting a portion of said image that covers at least said mouth neighborhood into an edge map;

an edge information calculator, for calculating edge information for said mouth neighborhood based on said edge map outputted by said converter;

and that said classifier classifies said candidate for human face region based on said edge information calculated by said edge information calculator.

18. The apparatus for processing an image according to claim 17, characterized in that said edge information calculator comprises:

a bright area identifier, for identifying within said edge map a series of bright areas each of which is composed of pixels whose characteristic values are greater than a first threshold;

a size calculator, for calculating the size of said candidate for human face S1 and the size of the biggest bright area S2;

and in that said classifier classifies said candidate for human face region as a false human face, or one candidate with high possibility of being a false human face if S2/S1 is less than a second threshold.

19. The apparatus for processing an image according to claim 18, characterized in that said size calculator also calculates the size of the second biggest bright area S3;

and in that said classifier further classifies said candidate for human face region as a false human face, or one candidate with high possibility of being a false human face if S3/S2 is less than a third threshold and the center of the biggest bright area is not in a mouth area, said mouth area being a predefined portion of said mouth neighborhood.

20. The apparatus for processing an image according to claim 19, characterized in that said classifier further classifies said candidate for human face region as a real human face, or one candidate with high possibility of being a real human face if S3/S2 is less than said third threshold and said center of said biggest bright area is in said mouth area.

21. The apparatus for processing an image according to claim 18, characterized in that said size calculator also calculates the size of the second biggest bright area S3;

and in that said classifier further classifies said candidate for human face region as a false human face, or one candidate with high possibility of being a false human face if S3/S2 is not less than a third threshold and the mid-point between the centers of the first two biggest bright areas is not in a mouth area, said mouth area being a predefined portion of said mouth neighborhood.

22. The apparatus for processing an image according to claim 21, characterized in that said classifier further classifies said candidate for human face region as a real human face, or one candidate with high possibility of being a real human face if S3/S2 is not less than said third threshold and said mid-point between said centers of said first two biggest bright areas is in said mouth area.

23. The apparatus for processing an image according to claim 17, characterized in that said edge information calculator comprises:

an edge intensity calculator, for calculating the average edge intensity within said mouth neighborhood and the average edge intensity within a mouth area, said mouth area being a predefined portion of said mouth neighborhood;

and in that said classifier classifies said candidate for human face region as a false human face, or one candidate with high possibility of being a false human face if the difference between said average edge intensity within said mouth area and said average edge intensity within said mouth neighborhood is not greater than a fourth threshold.

24. The apparatus for processing an image according to claim 23, characterized in that said classifier further classifies said candidate for human face region as a real human face, or one candidate with high possibility of being a real human face if said difference between said average edge intensity within said mouth area and said average edge intensity within said mouth neighborhood is greater than said fourth threshold.

25. The apparatus for processing an image according to any one of claims 19 to 22 and 23 to 24, characterized in that said mouth area is r2 times the width of said candidate for human face region, and at most r1 times the height of said candidate for human face region, wherein r2 is a constant greater than 0.1 and less than 0.7, and r1 is a constant greater than 0.1 and less than 0.7.

26. The apparatus for processing an image according to claim 25, characterized in that said r2 equals to 0.5, and said r1 equals to 0.3.