摘要： 提出一种基于人脸肤色统计模型和主元分析(PCA)的人脸检测和定位方法。对现有的高斯肤色模型进行改进，并提出一种新的简便有效的候选人脸区域定位方法，最后采用基于主元分析的人眼检测算法完成对候选区域的检验。实验证明该方法可以有效地运用于多人脸,不同尺寸和复杂背景的情况，具有良好的检测效果。

关键词: 人脸检测, 肤色模型, 主元分析, 特征眼

Abstract: This paper presents a human face detection and localization approach which is based on skin color detection and principle component analysis(PCA) . The combined gauss model(CGM) derived of the present gauss model is used for the skin color model. Additionally, a new simple and effective algorithm is presented, increasingly decreasing the time of the following eye-detection operation. To overcome the problem of skin color ambiguity, an eye detection algorithm based on PCA is used for the identification of face.

中图分类号: 

• TP391.41