摘要： 提出了一种基于肤色信息的人脸检测方法。构造了皮肤颜色模型、检测图像的皮肤区域，并计算出该区域相对于皮肤的马氏距离图。根据马氏距离图寻找特征脸子空间，构造出BP神经网络。以马氏距离图在各特征脸子空间的投影系数为BP神经网络的输入，计算该区域是否为人脸图像。实验证明，该方法是有效的。

关键词: 人脸检测, 马氏距离图, 特征脸, BP神经网络

Abstract: A method of face detection is proposed, which is based on the color of skin. A skin color model is constructed, which detects skin area in a color image and figures out map of Mahalanobis distance to skin. Based on map of Mahalanobis distance eigenface subspace is found out and a BP neural network model is constructed. On the BP neural network, input is coefficient vector obtained from map of Mahalanobis distance projecting on subspace, and output determines whether this area is face image. Experimental results show the method is effective．

Key words: Face detection, Map of Mahalanobis distance, Eigenface, BP neural network

中图分类号: 

• TP391.41