说明：   基于肤色的人脸检测方法，包括肤色模型建立和肤色分割过程两大部分内容。在比较了颜色空间和两种模型后，选择了在 YCbCr 空间建立高斯模型。输入图像经过相似度计算、阈值分割、最后得到人脸区域。
(Based on the color of the face detection methods, including color, model and color segmentation process most of the contents of the two. In comparing the color space and the two models, the choice in the YCbCr space to establish Gaussian model. After the input image similarity computation, threshold segmentation, and finally get face region.)