关键词: YCgCr颜色空间, 图像分割, 人脸检测, 高斯模型

Abstract: Due to the clustering is excellent in a color space YCgCr.This paper provides a way of face detection in a color space YCgCr.Firstly，skin colors are projected in a color space CgCr and the fitting of distribution is carried through in order to wipe off a part of non skin color and gain the intersected image for the first time.Experimental results indicate that this fitting of distribution can reduce the great mass of processing pixels of non-skin color.Then this paper puts Color Space YCgCr as the color model to build a color space and use Gauss model to detect human face on a base of the intersected image.Experimental results indicate that Gauss model to detect human face on a base of fitting of distribution in CgCr color space can improve the speed of face detection and have upper detecting ratio.This way also has upper detection ratio and lower ratio of missing detection and low ratio of false detection.

Key words: YCgCr color space, image segmentation, face detection, Gauss model