摘要
基于AdaBoost算法训练了分类器,并编程实现了一个人脸检测系统,将其应用于驾驶员疲劳检测系统中,对人脸区域进行检测和跟踪。为了提高系统的鲁棒性和检测的速度与精度,依次应用曝光补偿,积分图等方法,使得系统具有较高的高检测精度和较短的检测时间,可以适应驾驶员疲劳检测的需要。
The classifier is trained based on Adaboost algorithm, and a face detection system has been realized as a part of drivers＇ fatigue inspection system, which is used to inspect and track faces. For better robustness and speed, such methods as light compensation and integral image algorithm were applied. Consequently, the detection accuracy was improved and the detection time was reduced. This system can meet the requirement of drivers＇ fatigue inspection.
出处
《交通与计算机》
2008年第4期140-143,共4页
Computer and Communications
基金
"十一五"国家科技支撑计划重点项目(批准号:2006BAJ18B01)资助
