Automatic Detection of Vertical Gaze Using an Embedded Imaging Device
A method of detecting and applying a vertical gaze direction of a face within a digital image includes analyzing one or both eyes of a face within an acquired image, including determining a degree of coverage of an eye ball by an eye lid within the digital image. Based on the determined degree of coverage of the eye ball by the eye lid, an approximate direction of vertical eye gaze is determined. A further action is selected based on the determined approximate direction of vertical eye gaze.
1 . A method for detecting and applying a face angle direction within a digital image, comprising:
acquiring and detecting a digital image including at least part of a face;
analyzing the face within the digital image, including determining an approximate angular offset of the face from normal;
based on the determined approximate angular offset from normal of the face, determining that the approximate face angle direction does not to lie within a directional range of a digital device, and
initiating a power save routine based on the determined approximate face angle direction that does not lie within said directional range of said digital device.
2 . The method of claim 1 , wherein the determining an approximate angular offset of the face from normal comprises applying a 3D face modeling method.
3 . The method of claim 1 , wherein the determining an approximate angular offset of the face from normal comprises applying a 2D rotational technique.
4 . The method of claim 1 , wherein the determining an approximate angular offset of the face from normal comprises applying an affine transformation technique.
5 . The method of claim 1 , wherein the determining an approximate angular offset of the face from normal comprises face orientation detection.
6 . The method of claim 5 , wherein the orientation detection comprises analyzing the relative locations of different facial objects.
7 . The method of claim 6 , wherein the different facial objects comprise eyes and mouth.
8 . The method of claim 5 , wherein the face orientation detection is implemented as part of a digital display device.
9 . The method of claim 5 , wherein the face orientation detection comprises determining whether the face orientation is in the plane or out of the plane.
10 . The method of claim 1 , wherein the determining an approximate angular offset of the face from normal comprises applying models for inclined faces or faces in profile, or both.
11 . The method of claim 1 , wherein the determining an approximate angular offset of the face from normal comprises applying models for faces looking up or looking down, or both.
12 . The method of claim 1 , wherein the determining an approximate angular offset of the face from normal comprises applying models for faces looking left or looking right, or both.
13 . A method for detecting and applying a facial pose within a digital image, comprising:
acquiring and detecting a digital image including at least part of a face;
analyzing the face within the digital image, including determining a facial pose of the face;
based on the determined facial pose of the face, determining that the approximate direction of the facial pose does not to lie within an approximate directional range of a digital device; and
initiating a power save routine based on the determined facial pose that does not lie within said approximate directional range of said digital device.
14 . The method of claim 13 , wherein the determining that the approximate direction of the facial pose does not to lie within an approximate directional range of a digital device comprises determining that the approximate direction varies by more than 5% from an averaged pose.
15 . The method of claim 13 , wherein the determining an approximate angular offset of the face from normal comprises applying a 3D face modeling method.