| 基于Boosting算法的实时人脸监控系统设计 |
| |
| 引用本文： | 孙双宁,孙雁飞.基于Boosting算法的实时人脸监控系统设计[J].信息技术,2009,33(1). |
| |
| 作者姓名： | 孙双宁 孙雁飞 |
| |
| 作者单位： | 1. 吉林大学珠海学院电子信息科学与技术系,珠海,519041
2. 上海交通大学电子工程系图像通信与信息处理研究所,上海,200240 |
| |
| 摘 要： | 设计实现了基于Boosting算法的实时人脸监控系统,以AdaBoost算法为人脸检测基础,以粒子滤波器算法为人脸跟踪基础,通过两者的结合提高了检测的速度.通过分析AdaBoost训练和检测的过程,指出影响AdaBoost检测速度的要素,并提出了通过区域生长等预处理方式对待检测图像进行区域合并,降低背景的复杂度,从而提高检测的速度;并增加了侧面人脸级联分类器,采用串并联结构将正面人脸和侧面人脸的检测综合起来,扩大了系统对人脸的检测范围.同时将跟踪结果作为人脸检测模块的反馈信号,增强了检测系统的目标捕获和目标校正能力.
|
| 关 键 词： | 人脸监控 串并联结构 反馈校验 AdaBoost算法 |
Design of real-time face surveillance system based on Boosting algorithm |
| |
| SUN Shuang-ning,SUN Yan-fei.Design of real-time face surveillance system based on Boosting algorithm[J].Information Technology,2009,33(1). |
| |
| Authors: | SUN Shuang-ning SUN Yan-fei |
| |
| Affiliation: | 1.Department of Electronic Information Science & Technology;Zhuhai College;Jilin University;Zhuhai 519041;China;2.Institute of Image Communication and Information Processing;Department of Electronic Engineering;Shanghai Jiaotong University;Shanghai 200240;China |
| |
| Abstract: | This paper introduced a method of face surveillance based on Boosting algorithm,combined face detection with face tracking so as to promote the processing speed.Through analyzing the training and detecting algorithm of AdaBoost, this paper pointed out the reasons affecting the speed of AdaBoost mostly,and carried out a way of speeding up by region growth to reduce the complexity of background.Besides,profiles were added and united with frontal faces in a framework mixed with serial and parallel links to wid... |
| |
| Keywords: | face surveillance serial and parallel structure feedback checkup AdaBoost algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
