摘要
提出了一种结合人脸检测和跟踪的快速动态人脸检测算法。在层次型AdaBoost检测算法的基础上,结合提出的一种基于线性预测的人脸跟踪算法来缩小搜索空间,使得检测速度提高到24ms/帧。还设计了一种结合层次型AdaBoost算法、负样本反馈方式和利用权值缩减训练时间的检测器训练方法,可以方便地应用于物体检测技术上。
A fast face detection algorithm in video integrated with tracking is designed. It uses a cascade-structured face detector trained with AdaBoost to detect face. Integrated with a linear face tracker, searching region and rescaling levels of the detector are reduced, and thus detection speed is accelerated. It also designs a classifier training method integrated with AdaBoost, bootstrapping and weight-trimming, which can be easily used in object detection.
出处
《计算机工程》
CAS
CSCD
北大核心
2006年第11期222-224,共3页
Computer Engineering
基金
国家"十五"科技攻关计划重点资助项目"防范
打击重大刑事犯罪关键技术研究"(2001BA801B07)
