| 摘要 |
In the present disclosure, a plurality of frames of input images sequentially received for a predetermined time interval is obtained, and a face detecting operation is performed on a first frame if a full detecting mode is implemented. If a face is detected from a specific region of the first frame during the face detecting operation, a face tracking mode is implemented, a second frame is divided to produce the divided input image portions of the second frame, and the face tracking operation is performed on a surrounding region of the specific region of the divided input image portions of the second frame that corresponds to the specific region in the first frame. If the face is not detected in the face tracking mode, a partial detecting mode is implemented, and the face detecting operation is performed on image portions resized on divided input image portions of a third frame to which a specific region of the third frame corresponding to the specific region of the first frame belongs. |
