基于VC++.net和OPENCV的身份识别系统开发

The Development of Identity Recognition System Based on VC++.net and OPENCV

首先使用Adaboost方法检测到视频流中的人脸,然后将SIFT特征提取算法应用到人脸特征的提取环节中,提取出人脸的旋转、缩放、平移和光照不变性SIFT特征。通过将检测到的人脸的SIFT特征和人脸库中的人脸SIFT特征进行匹配,找出匹配程度最大的对应的类别,即为匹配结果。整个系统用VisualC++.net2008和OPENCV开发。

First,this paper used Adaboost to detect the face in the video, and then used the method of SIFT which with the invariant of rotation, scale, translation and illumination to extract the feature of face. Practice illustrated that the same face has the same SIFT, by matching the SIFT, we can recognize the face. The identity recognition system is developed by Visual C++.net 2008 and OPENCV(open source computer vision).