Facial feature detecting apparatus and facial feature detecting method

1. A facial feature detecting apparatus comprising:

a feature sensor configured to acquire information on facial features of a subject;

a feature detecting unit configured to detect the facial features including a first facial feature and a second facial feature of the subject from the information acquired by the feature sensor and determine whether the first facial feature and the second facial feature are detected;

a three-dimensional coordinates calculating unit configured to calculate three-dimensional coordinates of the facial features of the subject; and

a feature position estimating unit configured to estimate first three-dimensional coordinates of the first facial feature from second three-dimensional coordinates of the second facial feature of the detected facial features, on a basis that the first facial feature and the second facial feature are located at bilaterally symmetrical positions, wherein the facial feature detecting apparatus further comprises:

a median line estimating unit configured to estimate a median line of a face of the subject, based on the three-dimensional coordinates of the detected facial features if the first facial feature is not detected and the second facial feature is detected,

wherein, the feature position estimating unit estimates, as the first three-dimensional coordinates of the first facial feature, a position that is symmetrical to a position of the second facial feature of the detected facial features with respect to the median line, and

wherein the facial feature detecting apparatus further comprises a 3D model accumulating unit configured to calculate an average of the second three-dimensional coordinates of the facial features accumulated in a 3D model storage at regular intervals for each facial feature type, and delete three-dimensional coordinates of a facial feature that is farthest from the average, from the 3D model storage.

2. The facial feature detecting apparatus according to claim 1 ,

wherein the first and the second facial features are ears of the subject, and

wherein the facial feature detecting apparatus further comprises an active noise control controller configured to perform an active noise control upon receiving the estimated first three-dimensional coordinates of the first facial feature and the detected second three-dimensional coordinates of the second facial feature.

3. A facial feature detecting apparatus comprising:

a feature sensor configured to acquire information on facial features of a subject;

a feature detecting unit configured to detect the facial features of the subject from the information acquired by the feature sensor;

a three-dimensional coordinates calculating unit configured to calculate first three-dimensional coordinates of the facial features of the subject;

a face direction estimating unit configured to estimate a face direction of the subject, based on the calculated first three-dimensional coordinates of the detected facial features;

a 3D model information acquiring unit configured to acquire given three-dimensional coordinates of a given facial feature of the subject from a 3 D model storage that accumulates second three-dimensional coordinates of the facial features of the subject in a frontal face direction; and

a face direction rotating unit configured to cause the acquired given three-dimensional coordinates of the given facial feature of the subject to be rotated in the estimated face direction,

wherein the facial feature detecting apparatus further comprises:

a 3D model accumulating unit configured to accumulate, in the 3D model storage, the calculated second three-dimensional coordinates of the facial features of the subject in the frontal face direction, and

wherein the 3D model accumulating unit calculates an average of the second three-dimensional coordinates of the facial features accumulated in the 3D model storage at regular intervals for each facial feature type, and deletes three-dimensional coordinates of a facial feature that is farthest from the average, from the 3D model storage.

4. The facial feature detecting apparatus according to claim 3 , wherein, when the feature detecting unit is unable to detect the given facial feature of the subject from the information acquired by the feature sensor, the 3D model information acquiring unit acquires the given three-dimensional coordinates of the given facial feature of the subject from the 3D model storage.

5. The facial feature detecting apparatus according to claim 3 , further comprising:

a face direction reverse rotation unit configured to cause the facial features of the subject to be reversely rotated in accordance with the estimated face direction, and calculate the second three-dimensional coordinates of the facial features of the subject in the frontal face direction.

6. The facial feature detecting apparatus according to claim 5 , wherein, when the estimated face direction is the frontal face direction, the 3D model accumulating unit does not accumulate, in the 3D model storage, second three-dimensional coordinates of an ear of the subject in the frontal face direction.

7. A facial feature detecting method comprising:

acquiring, by a feature sensor, information on facial features of a subject;

detecting, by a feature detecting unit, the facial features of the subject from the information acquired by the feature sensor;

calculating, by a three-dimensional coordinates calculating unit, first three-dimensional coordinates of the facial features of the subject;

estimating, by a face direction estimating unit, a face direction of the subject, based on the calculated first three-dimensional coordinates of the detected facial features;

acquiring, by a 3D model information acquiring unit, given three-dimensional coordinates of a given facial feature of the subject from a 3D model storage that accumulates second three-dimensional coordinates of the facial features of the subject in a frontal face direction; and

causing, by a face direction rotating unit, the acquired given three-dimensional coordinates of the given facial feature of the subject to be rotated in the estimated face direction,

wherein the facial feature detecting method further comprises:

accumulating, in the 3D model storage, the calculated second three-dimensional coordinates of the facial features of the subject in the frontal face direction,

calculating an average of the second three-dimensional coordinates of the facial features accumulated in the 3D model storage at regular intervals for each facial feature type, and deleting three-dimensional coordinates of a facial feature that is farthest from the average, from the 3D model storage, and

calculating an average of the second three-dimensional coordinates of the facial features accumulated in the 3D model storage at regular intervals for each facial feature type, and deleting three-dimensional coordinates of a facial feature that is farthest from the average, from the 3D model storage.