基于YCbCr肤色模型和SNOW分类器的人脸检测

Face Detection Based on YCbcr Skin Color Model and SNOW Classifier

人脸检测是人脸识别的首要步骤,在人脸识别领域有重要的应用价值。基于YCbCr彩色空间,提出一种RGB彩色图像的人脸检测方法。该方法利用YCbCr肤色模型进行肤色分割,得到类肤色区域作为侯选人脸区域;结合split up Sparse Network OfWinnows(SNOW)分类器准确定位人脸的位置应用matlab编程技术对多组图像进行实验,结果表明,该方法适用于复杂条件下的人脸检测,并且不受人脸表情的限制,对于多人脸检测同样适用。

Face detection,the first step of face recognition,is significant to application.A novel method for face detection in color images is presented by using YCbCr color space.The skin color segmentation is carried out and this skin-lile region is used to be candidate face.Sparse Network Of Winnows（SNOW）classifier is used to detect face region precisely.With Matlab Programming used in the experiment,the results show that the face detection method,which isn＇t limited by expression,is suitable in comples background and in detecting many faces.