Data processing apparatus, data processing method and recording medium

• US 7,916,977 B2
• Filed: 12/23/2009
• Issued: 03/29/2011
• Est. Priority Date: 03/12/1999
• Status: Expired due to Fees

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Ishibashi, Junichi, Kondo, Tetsujiro
• Primary Examiner(s): Le; Brian Q
• Application Number:
  US12/645,643
• Publication Number:
  US 20100098344A1
• Time in Patent Office: 461 Days
• Field of Search: 382/115, 382/118, 382/293, 382/294, 382/295, 382/296, 382/297, 382/305, 707/736, 707/737, 707/791, 707/802
• US Class Current: 382/305
• CPC Class Codes: G06T 2207/30201 Face
• CPC Class Codes: G06T 7/74 involving reference images or patches
• CPC Class Codes: G06V 40/165 using facial parts and geometric relationships

Abstract

A group of input image data is determined based on the distance between images of a number of representative faces, prepared from plural templates, and the input image data (correlative values C1 to C5). The number of representative faces is smaller than the number of the templates. An angle of the input image is calculated based on the so-determined group.

Claims

1. An image processing method for building a database of facial orientation image templates, comprising:

generating a plurality of facial images oriented in different directions;

providing a range for each image of said plurality of facial images;

determining an orientation angle for said each image by detecting and calculating distances among predetermined features within the range;

linking the determined orientation angle to said each image; and

storing said plurality of facial images linked to orientation angles as facial orientation image templates in the database.

2. The image processing method of claim 1, wherein said providing a range comprises

extracting a rectangle including both eyes and a nose.

3. The image processing method of claim 2, further comprising specifying positions of both eyes and the nose.

4. The image processing method of claim 3, wherein said determining an orientation angle comprises

calculating a center of gravity of an area specified by the positions of both eyes and the nose by entering the area within the rectangle.

5. The image processing method of claim 1, further comprising classifying the plurality of facial images into a plurality of face classes depending on orientation angles.

6. The image processing method of claim 5, wherein said classifying the plurality of facial images comprises

classifying the plurality of facial images into three face classes, oriented toward front, right, and left.

7. The image processing method of claim 5, further comprising:

calculating a representative facial image for each class of said plurality of face classes by calculating a mean value of facial images in each class.

8. The image processing method of claim 7, wherein said calculating a representative facial image for each class comprises

averaging each facial image by registering the positions of both eyes according to several facial orientation image templates to account for differences in effective areas among said plurality of face classes.

9. The image processing method of claim 7, wherein said calculating a representative facial image for each class comprises

calculating a correlative value for the representative facial image.

10. The image processing method of claim 9, further comprising classifying the representative facial image into one of said plurality of face classes, where said one of said plurality of face classes includes facial images having the correlative value above a predetermined threshold value.