基于码本（codebook）的背景建模的背景差分法+级联基于LBK或haar的adaboost和基于hog的svm分类器+快速hough圆变换进行人头识别+基于区域特征的目标跟踪算法。(编程)
AdaBoost是一种增强性机器学习算法，它用于把弱分类器联合成强分类分类器(Background modeling based on codebook (codebook) background difference method + cascade based on LBK or Haar AdaBoost and hog based SVM Classifier + fast Hough circle transform for human head recognition + target tracking algorithm based on regional characteristics.)