Face detection

• US 7,336,830 B2
• Filed: 11/25/2003
• Issued: 02/26/2008
• Est. Priority Date: 11/29/2002
• Status: Expired due to Fees

• Current Assignee: Sony United Kingdom Limited (Sony Group Corp.)
• Original Assignee: Sony United Kingdom Limited (Sony Group Corp.)
• Inventors: Porter, Robert Mark Stefan, Rambaruth, Ratna
• Primary Examiner(s): MARIAM, DANIEL G
• Application Number:
  US10/723,093
• Publication Number:
  US 20040151381A1
• Time in Patent Office: 1,554 Days
• Field of Search: 382/103, 382/118, 382/165, 382/170, 382/209, 382/218, 382/282, 382/288, 340/5.53, 358/538
• US Class Current: 382/218
• CPC Class Codes: G06V 10/48 by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation
• CPC Class Codes: G06V 10/507 Summing image-intensity values; Histogram projection analysis
• CPC Class Codes: G06V 40/161 Detection; Localisation; Normalisation
• CPC Class Codes: H04N 23/611 where the recognised objects include parts of the human body
• CPC Class Codes: H04N 5/772 the recording apparatus and the television camera being placed in the same enclosure

Abstract

Video face detection apparatus in which a test image from a video sequence is compared with an image property model derived from image properties of a region detected to contain a face in a preceding image in the video sequence comprises: selecting a predetermined proportion of pixels in the region detected to contain a face in the preceding image which most closely match the image property model derived in respect of that region, thereby deriving a pixel mask; and comparing pixels in the test image defined by the pixel mask with the image property model, the mask being applied at more than one image position within the test image; a face being detected in the test image at a mask position corresponding to a lowest average difference between the image property model and pixels defined by the mask at that position.

Claims

1. A video face detection apparatus in which a test image from a video sequence is compared with an image property model derived from image properties of a region detected to contain a face in a preceding image in said video sequence, said apparatus comprising:

(i) a selector to select a predetermined proportion of pixels in said region detected to contain a face in said preceding image which most closely match said image property model derived in respect of that region, thereby deriving a pixel mask;

(ii) a comparator to compare pixels in said test image defined by said pixel mask with said image property model, said mask being applied at more than one image position within said test image;

a face being detected in said test image at a mask position corresponding to a lowest average difference between said image property model and pixels defined by said mask at that position.

2. The apparatus according to claim 1, in which said image property model is a color model.

3. The apparatus according to claim 2, in which said color model is a Gaussian model of color distribution.

4. The apparatus according to claim 2, in which said color model represents a color distribution in at least a part of at least one image of said video sequence.

5. The apparatus according to claim 1, in which said mask is applied to said test image at positions within a test region surrounding the image position of said detected face in said preceding image.

6. The apparatus according to claim 5, in which said test region is a rectangular region.

7. The apparatus according to claim 1, in which said predetermined proportion is 50% of said pixels.

8. A video conferencing apparatus comprising the apparatus according to claim 1.

9. A surveillance apparatus comprising the apparatus according to claim 1.

10. A video face detection method in which a test image from a video sequence is compared with an image property model derived from image properties of a region detected to contain a face in a preceding image in said video sequence, said method comprising the steps of:

(i) selecting a predetermined proportion of pixels in said region detected to contain said face in said preceding image which most closely match said image property model derived in respect of that region, thereby deriving a pixel mask; and

(ii) comparing pixels in said test image defined by said pixel mask with said image property model, said mask being applied at more than one image position within said test image;

said face being detected in said test image at a mask position corresponding to a lowest average difference between said image property model and pixels defined by said mask at that position.

11. A computer readable medium encoded with a computer program configured to cause a processor to execute a video face detection method in which a test image from a video sequence is compared with an image property model derived from image properties of a region detected to contain a face in a preceding image in said video sequence, said method comprising the steps of:

(i) selecting a predetermined proportion of pixels in said region detected to contain said face in said preceding image which most closely match said image property model derived in respect of that region, thereby deriving a pixel mask; and

(ii) comparing pixels in said test image defined by said pixel mask with said image property model, said mask being applied at more than one image position within said test image;

said face being detected in said test image at a mask position corresponding to a lowest average difference between said image property model and pixels defined by said mask at that position.