基于Adaboost算法的行人检测

Pedestrians Detection Based on Adaboost Algorithm

提出一种基于Adaboost算法的行人检测方法。Adaboost是将一组弱分类器通过一定的规则,结合成为一个强分类器,再把这些强分类器级联成为一个快速、准确的分类器。实验证明基于此算法的行人检测具有检测率高、速度快的特点,能够达到实时检测的要求。

This paper presents a method based on Adaboost algorithm for pedestrians detection.Adaboost algorithm can con-struct a strong classifier by combining a series of weak classifiers through some rules fastly and accurately,then cascaded classifier could be obtained by cascading existing strong classifiers.The method of using Adaboost algorithm in vehicle de-tection has faster speed and higher correctness than most of the existing methods in the experiments.