中文摘要：

基于视频的自动人脸识别系统自动选取用于后续识别的最佳人脸图像.设计并实现了一个最佳人脸捕获模块,该模块主要由人脸检测、人脸跟踪和姿态判定环节组成.在实现对视频中的人脸进行检测和跟踪的基础上,可自动截取并保存一张大小合适、清晰度高且姿态最佳的人脸,用于后续的人脸识别.

英文摘要：

Different from the video-based face recognition systems which needs select faces manually, video-based automatic face recognition systems should be able to automatically pick out an optimal faces for subsequent face recognition. Therefore, a capture module for best face is designed and implemented. The module mainly consists of face detecting, face tracking and attitude determination links. Based on the detection and tracking of human face in video, an optimal face with appropriate size, high definition and best attitude can be captured automatically for the face recognition.