IMAGE PROCESSING APPARATUS, IMAGE PROCESSING METHOD, AND STORAGE MEDIUM

• US 20110164816A1
• Filed: 12/22/2010
• Published: 07/07/2011
• Est. Priority Date: 01/05/2010
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Guo, Muling
• Granted Patent: US 8,463,029 B2
• Time in Patent Office: Days
• US Class Current: 382/165
• CPC Class Codes: G06V 40/164 using holistic features

Abstract

According to one aspect of the invention, an image processing apparatus comprises: a face detection unit configured to detect a face area of a person from an image; a head detection area setting unit configured to set a head detection area based on the detected face area of the person; an edge detection unit configured to detect an edge from the set head detection area, and generate an edge image which is comprised of the detected edge; an edge deletion unit configured to delete an edge existing between the face and hair from the edge image; and an ellipse detection unit configured to detect a candidate ellipse corresponding to a head area from an edge image from which an edge is deleted by the edge deletion unit.

Claims

1. An image processing apparatus comprising:

a face detection unit configured to detect a face area of a person from an image;

a head detection area setting unit configured to set a head detection area based on the detected face area of the person;

an edge detection unit configured to detect an edge from the set head detection area, and generate an edge image which is comprised of the detected edge;

an edge deletion unit configured to delete an edge existing between the face and hair from the edge image; and

an ellipse detection unit configured to detect a candidate ellipse corresponding to a head area from an edge image from which an edge is deleted by said edge deletion unit.

2. The apparatus according to claim 1, wherein in accordance with a predetermined criterion, said ellipse detection unit selects one candidate ellipse from a plurality of candidate ellipses obtained by performing Hough transformation for the edge image from which the edge is deleted, and sets the selected candidate ellipse as a head area.

3. The apparatus according to claim 1, wherein said edge deletion unit is further comprising:

a skin color area detection unit configured to detect a skin color area from the head detection area based on a color of the face area; and

an expansion unit configured to expand the skin color area and delete an edge existing in the expanded skin color area.

4. The apparatus according to claim 3, wherein said skin color detection unit comprises:

a unit configured to obtain a color distribution of pixels in the face area;

a unit configured to obtain a skin color distribution by deleting a predetermined portion from the color distribution; and

a unit configured to detect a skin color area from the head detection area by using the skin color distribution.

5. The apparatus according to claim 1, wherein said edge deletion unit comprises:

a skin color area detection unit configured to detect a skin color area from the head detection area based on a color of the face area;

a hair color sample pixel detection unit configured to detect a hair color sample pixel from a hair sample candidate area having a predetermined positional relationship with the skin color area detected by said skin color area detection unit;

a unit configured to obtain a color distribution of the hair color sample pixels;

a unit configured to detect a hair color area from the head detection area by using a color distribution range of the hair color sample pixels; and

a unit configured to delete an edge pixel between the skin color area and the hair color area.

6. The apparatus according to claim 5, wherein said hair color sample pixel detection unit comprises:

a unit configured to set the hair sample candidate area having a predetermined positional relationship with the skin color area;

a unit configured to generate a hair sample pixel candidate by removing a skin color pixel from the hair sample candidate area; and

a unit configured to detect a pixel located above the skin color area from the hair sample pixel candidates and sets the detected pixel as a hair sample pixel.

7. The apparatus according to claim 5, wherein said edge deletion unit further comprises:

a unit configured to expand the skin color area;

a unit configured to expand the hair color area; and

a unit configured to remove an edge pixel in an overlapping area between the expanded skin color area and the expanded hair color area.

8. The apparatus according to claim 5, wherein said unit configured to delete an edge pixel between the skin color area and the hair color area is further configured to delete the edge pixel between a lowermost pixel in the hair color area and an uppermost pixel in the skin color area, in each column in the skin color area, when a distance between the lowermost pixel in the hair color area and the uppermost pixel in the skin color area falls within a predetermined range.

9. The apparatus according to claim 1, further comprising a setting unit configured to manually set whether said edge deletion unit is caused to operate or not.

10. The apparatus according to claim 1, further comprising a validity determining unit configured to determine, based on a determination criterion, whether the candidate ellipse detected by said ellipse detection unit is valid,

wherein when said validity determining unit determines that the candidate ellipse is valid, said edge deletion unit is operated.

11. The apparatus according to claim 10, wherein said validity determining unit comprises at least one of units including:

a unit configured to determine whether the candidate ellipse overlaps a face area or a portion of the face area;

a unit configured to determine whether an edge pixel count on an elliptic boundary of the candidate ellipse is not less than a predetermined threshold;

a unit configured to determine whether a ratio at which edge pixels on an elliptic arc extending from a leftmost point of the edge pixels to a rightmost point of the edge pixels cover edge pixels on the elliptic boundary of the candidate ellipse is not less than a predetermined threshold;

a unit configured to determine whether a ratio at which the edge pixels on the elliptic boundary of the candidate ellipse cover an elliptic circumference of the head area is not less than a predetermined threshold;

a unit configured to determine whether an aspect ratio of the candidate ellipse falls within a predetermined range; and

a unit configured to determine whether a ratio of a skin color area in the candidate ellipse is not less than a predetermined threshold.

12. An image processing method comprising steps of:

detecting a face area of a person from an image;

setting a head detection area based on the detected face area of the person;

detecting an edge from the set head detection area and generating an edge image which is comprised of the detected edge;

deleting an edge existing between the face and hair from the edge image; and

detecting a candidate ellipse corresponding to a head area from an edge image from which an edge is deleted in said deleting step.

13. A storage medium storing a computer program for causing a computer system to execute steps defined in claim 12.