Distinguishing true 3-d faces from 2-d face pictures in face recognition

1. An image processing device, comprising:

an obtaining unit configured to obtain a plurality of images captured in time series;

a first calculating unit configured to calculate a first change vector indicating a change between the images in an angle representing a posture of a subject included in each of the images;

a second calculating unit configured to calculate a second change vector indicating a change in coordinates of a feature point of the subject;

a third calculating unit configured to calculate an intervector angle between the first change vector and the second change vector; and

a determining unit configured to determine that the subject is three-dimensional when the intervector angle is smaller than a predetermined first threshold.

2. The device according to claim 1 , wherein

the subject is a human face,

the angle representing the posture is a face orientation angle obtained based on a rotation matrix obtained by multiplying a predetermined pseudo inverse matrix of a matrix composed of three-dimensional coordinates of the feature point by a matrix composed of coordinates of the feature point in the images,

the first calculating unit calculates the first change vector that is a two-dimensional vector remaining as a result of excluding one component in an image plane from a three-dimensional rotational motion that is a temporal change of the rotation matrix, and

the second calculating unit calculates the second change vector indicating a change in the feature point that is at least one of a left or right nostril, a midpoint of left and right nostrils and a tip of a nose.

3. The device according to claim 2 , further comprising:

a frame information storing unit configured to store therein the face orientation angle and the feature point; and

a frame evaluation value calculating unit configured to calculate an evaluation value of the images, wherein

the frame evaluation value calculating unit identifies the face orientation angle at which a magnitude of the first change vector calculated for an image to be processed is larger than a predetermined magnitude from a plurality of face orientation angles stored in the frame information storing unit, and identifies an image associated with the identified face orientation angle from the plurality of images, and

the first calculating unit calculates the first change vector that is a two-dimensional vector remaining as a result of excluding one component in an image plane from a three-dimensional rotational motion that is a temporal change between the face orientation angle in the image to be processed and the face orientation angle in the identified image.

4. The device according to claim 1 , further comprising:

a detecting unit configured to detect coordinates of at least three feature points of the subject; and

a fourth calculating unit configured to calculate relative coordinates of coordinates of one of the three feature points based on coordinates of other two of the feature points, wherein

the subject is a human face,

the angle representing the posture is a face orientation angle obtained based on a rotation matrix obtained by multiplying a predetermined pseudo inverse matrix of a matrix composed of three-dimensional coordinates of the feature points by a matrix composed of coordinates of the feature points in the images, and

the determining unit determines that the subject included in the images is three-dimensional when a change in the relative coordinates of the feature points in the images with different face orientation angles is larger than a predetermined second threshold.

5. The device according to claim 4 , wherein

the detecting unit detects the coordinates of at least three feature points on a face center line of a human face that is the subject.

6. The device according to claim 4 , wherein

the second calculating unit calculates the second change vector indicating a change in the relative coordinates of the feature points of the subject between the plurality of images.

7. The device according to claim 4 , further comprising:

an evaluation value calculating unit configured to calculate an evaluation value by using a function to which a plurality of intervector angles at different times or different feature points are input, wherein

the determining unit determines whether or not the subject included in the images is three-dimensional by using the evaluation value.

8. An image processing method comprising:

obtaining a plurality of images captured in time series;

calculating a first change vector indicating a change between the images in an angle representing a posture of a subject included in each of the images;

calculating a second change vector indicating a change in coordinates of a feature point of the subject;

calculating an intervector angle between the first change vector and the second change vector; and

determining that the subject included in the images is three-dimensional when the intervector angle is smaller than a predetermined first threshold.

9. A computer program product comprising a nontransitory computer readable medium including programmed instructions, wherein the instructions, when executed by a computer, cause the computer to perform:

obtaining a plurality of images captured in time series;

calculating a first change vector indicating a change between the images in an angle representing a posture of a subject included in each of the images;

calculating a second change vector indicating a change in coordinates of a feature point of the subject;

calculating an intervector angle between the first change vector and the second change vector; and

determining that the subject included in the images is three-dimensional when the intervector angle is smaller than a predetermined first threshold.