Face location detection

Patent metadata

Field	Value
Publication number	US-9582706-B2
Application number	US-201214114213-A
Country	US
Kind code	B2
Filing date	Apr 23, 2012
Priority date	Apr 28, 2011
Publication date	Feb 28, 2017
Grant date	Feb 28, 2017

Abstract

The location of a face is detected from data about a scene. A 3D surface model from is obtained from measurements of the scene. A 2D angle data image is generated from the 3D surface model. The angle data image is generated for a virtual lighting direction, the image representing angles between a ray directions from a virtual light source direction and normal to the 3D surface. A 2D face location algorithm is applied to each of the respective 2D images. In an embodiment respective 2D angle data images for a plurality of virtual lighting directions are generated and face locations detected from the respective 2D images are fused.

First claim

The invention claimed is:

1. An image processing method wherein a location of a face is detected, the method comprising
obtaining a 3D surface model from measurements of a scene;
generating a 2D image of angle data from the 3D surface model, the 2D image representing angle data, the angle data for each respective image point in the 2D image being selected dependent on an angle between an incidence direction derived from a virtual lighting direction and a normal to the 3D surface at a point on the 3D surface that is in view in the 2D image at the image point;
applying a 2D face location algorithm to the 2D image
generating a plurality of respective 2D images from the 3D surface model, each representing the angle data for a respective virtual lighting direction;
applying the 2D face location algorithm to each of the respective 2D images;
combining face locations detected from the respective 2D images.

2. A method according to claim 1 , comprising generating a plurality of said respective 2D images for a same viewing direction for said respective virtual lighting directions.

3. A method according to claim 2 , comprising generating respective pluralities of said respective 2D images from, each plurality from a different viewing direction, each plurality comprising respective 2D images for a plurality of mutually different respective virtual lighting directions.

4. A method according to claim 2 , comprising determining an average detected 2D location for said plurality of respective 2D images and determining a 3D surface location associated with said average detected 2D location according to the 3D surface model.

5. A method according to claim 1 , comprising determining respective 3D surface locations associated with respective detected 2D face locations according to the 3D surface model and determining an average of the respective 3D surface locations.

6. A method according to claim 1 , wherein the 3D surface model is obtained using information from one or more further 2D images of light received from the scene.

7. A method according to claim 6 , comprising lighting the scene by means of structured light.

8. A method according to claim 6 , wherein the respective 2D images are generated for a viewing direction corresponding to a viewing direction of the one or more further 2D images of light received from the scene.

9. A method according to claim 6 , wherein the respective 2D images are generated independent of light reflection properties associated with the points on the 3D surface, if any.

10. A method according to claim 1 , comprising using a face location detected by the 2D face location algorithm and the 3D surface model to determine a 3D location of a point on the 3D surface that is visible at the detected face location.

11. A non-transitory computer program product, comprising a program of instruction for a programmable processor that, when executed by the programmable processor, will cause the programmable processor to perform the method of claim 1 .

12. An image processing method wherein a location of a face is detected, the method comprising
obtaining a 3D surface model from measurements of a scene;
generating a 2D image of angle data from the 3D surface model, the 2D image representing angle data, the angle data for each respective image point in the 2D image being selected dependent on an angle between an incidence direction derived from a virtual lighting direction and a normal to the 3D surface at a point on the 3D surface that is in view in the 2D image at the image point;
applying a 2D face location algorithm to the 2D image, wherein applying said 2D face location algorithm comprises determining a plurality of sums of the angles over respective regions in the image, comparing the sums with thresholds and combining the results of said comparisons.

13. An image processing system, the system comprising
a 2D image generator configured to generate a number of 2D images from a 3D surface model obtained for a scene, the 2D images representing angle data, the angle data for each respective image point in the 2D image being selected dependent on an angle between an incidence direction derived from a virtual lighting direction and a normal to the 3D surface at a point on the 3D surface that is in view in the image at the image point;
a 2D face location detector configured to apply a 2D face location detection algorithm to each of the respective 2D images, wherein the 2D image generator is configured to generate a plurality of respective 2D images from the 3D surface model, each respective 2D image representing the angle data for a respective virtual lighting direction, the image processing system comprising a data combiner configured to combine face location detection results detected by 2D face location detector from the plurality of respective 2D images.

14. An image processing system according to claim 13 , comprising a camera and a 3D model generator configured to generate the 3D surface model using image data from said camera.

15. An image processing system, the system comprising
a 2D image generator ( 32 ) configured to generate a number of 2D images from a 3D surface model obtained for a scene, the 2D images representing angle data, the angle data for each respective image point in the 2D image being selected dependent on an angle between an incidence direction derived from a virtual lighting direction and a normal to the 3D surface at a point on the 3D surface that is in view in the image at the image point;
a 2D face location detector ( 34 ) configured to apply a 2D face location detection algorithm to each of the respective 2D images by determining a plurality of sums of the angles over respective regions in the image, comparing the sums with thresholds and combining the results of said comparisons.

Assignees

• Van Bree Karl Catharina
• Vlutters Ruud
• Znamenskiy Dmitry Nikolayevich
• Koninklijke Philips Nv

Inventors

• Van Bree Karl Catharina
• Vlutters Ruud
• Znamenskiy Dmitry Nikolayevich

Classifications

• G06K9/00208 Primary

  Physics · mapped topic

• G06K9/00228

  Physics · mapped topic

• G06V40/161 Primary

  Detection; Localisation; Normalisation · CPC title

• G06V20/647 Primary

  by matching two-dimensional images to three-dimensional objects · CPC title

• G06T15/50

  Lighting effects · CPC title