Adaboost算法及其在实时视频流中的应用

针对Adaboost算法在实时视频流中的应用,本文基于Adaboost算法的人脸检测原理,即通过提取图像中的haar特征,在训练过程中选出最优特征,转换成弱分类器,优化组合于人脸检测。最终,利用opencv的开发包,通过VC++软件编程实现基于Adaboost算法实时视频流中的人脸检测。

According to the Adaboost algorithm in real-time video streaming application, based on the Adaboost algorithm for face detection principle, namely through the extraction of image features in gaar, portfolio the VC ＋ in the training process to choose the best feature, converted into a weak classifier, optimization in face detection. Finally, the use of Opener development kits, through ＋ software programming based on Adaboost algorithm for face detection in video stream.