本文的目的是借助Paul Viola最新提出的实时特征检测的技术，实现一个能够进行快速人脸检测的系统。并且通过训练，得到尽可能优化的分类器构造参数，从而获得具有高检测速度和检测正确率的人脸检测系统。-The purpose of this paper by drawing on the latest proposed by Paul Viola real-time feature detection technology, to realize a fast face detection system. Through training, to be as much as possible to optimize the classifier structure parameters to obtain a high detection speed and detection rate of correct face detection system.