快速人脸检测系统的设计与实现  

Design and implementation of fast face detection system

作　　者：吴暾华[1] 周昌乐[1] 

机构地区：[1]厦门大学计算机科学系,福建厦门361005

出　　处：《计算机应用》2005年第10期2351-2353,共3页journal of Computer Applications

摘　　要：为了快速而准确地在复杂背景的灰度或彩色图像中检测人脸,对实时人脸检测方法做了一些扩展并根据该方法设计实现了一个快速人脸检测系统,系统分为训练和检测两部分,训练方法为AdaBoost算法。测试表明该系统的性能较好。In order to detect faces in gray or color images at high speed and accuracy, some extension to the technology of real-time face detection were made and a rapid face detection system was designed and realized. The training subsystem is based on AdaBoost algorithm. The experiment results show that the extension to Viola＇ s technology is quite successful. The rapid face detection system is divided into two parts, training（ or learning） subsystem and detecting subsystem.

关 键 词：人脸检测 矩形特征 ADABOOST 层叠分类器 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]