实时视频中的快速人脸检测方法  

Fast Face Detection Method in Real-time Video

作　　者：索璐静[1] 陆小锋[1] 陆亨立[1] 张晶晶[1] 范天翔[1] 

机构地区：[1]上海大学通信与信息工程学院,上海200072

出　　处：《计算机工程》2011年第20期166-168,共3页Computer Engineering

基　　金：上海大学创新基金资助项目(A10-0107-09-902)

摘　　要：提出一种实时视频中的快速人脸检测方法。采用DirectShow获取和回放视频帧,进行前处理。对所得彩色图像进行肤色分割,通过肤色的统计特性得到人脸候选区域,将该区域设置为感兴趣区域。运用基于局部扫描的AdaBoost算法对每个感兴趣区域进行检测。实验结果证明了该方法的有效性。This paper proposes a fast face detection method in real-time video.It uses the DirectShow technology to grab the video frame,which has significantly increased system speed.After some pre-processes,the system detects the skin color by the statistical property of human skin to make the skin areas as the regions of interests.And these regions of interests are scanned by this improved AdaBoost algorithm for the further detection.Experimental results proves the validity of this method.

关 键 词：人脸检测 DIRECTSHOW技术 肤色分割 ADABOOST算法 局部扫描 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]