说明：  先用“gettrainingdata”获取人脸样本库的特征，样本库中每人只有一个样本。 然后用“facerecognition"输入待识别的人脸，可以较准确的得出结果！
(first "gettrainingdata" Access to face the sample characteristics, Each sample repository is only a sample. Then "facerecognition" to the importation of face recognition, the more accurate the results!)