基于粒特征和连续Adaboost的人脸检测  

Face detection using real Adaboost on granular features

作　　者：陈春燕[1] 章品正[1] 罗立民[1] 

机构地区：[1]东南大学影像科学与技术实验室,江苏南京210096

出　　处：《智能系统学报》2009年第5期446-452,共7页CAAI Transactions on Intelligent Systems

基　　金：国家自然科学基金资助项目(60702029)

摘　　要：提出了一种基于粒特征和连续Adaboost算法的人脸检测方法.它使用粒特征并扩展贝叶斯决策弱分类器,设计具有连续置信度输出的查找表型弱分类器形式,构造出弱分类空间,使用大规模的训练集和验证集,采用连续Adaboost算法学习得到Boosting动态级联型的人脸检测器.在CMU-MIT正面人脸测试集上,误报20个时,检测率为90%以上.在一台Pentium Dual-1.2 GHz的PC上,处理一幅大小为320×240像素大小的图片平均需100 ms.实验结果表明该方法取得了比较好的精度和速度.A face detection method based on sparse granular features and the real adaptive boosting （Adaboost） meta-algorithm was proposed. A sparse granular feature set was introduced into the Adaboost learning framework. A weak look-up-table （LUT） type classifier with real confidence output was designed by extending the Bayesian stump. Then, the space of the weak classifier was constructed. The Adaboost cascade face detector was taught by using a large training set and an evaluation set. Experiments were performed on the CMU-MIT dataset, a standard public data set for benchmarking frontal face detection systems. The detection rate reached over 90% when false alarms were 20. The average processing time on a Pentium Dual-1.2GHz PC was about 100 ms for a 320×240-pixel image. This shows the proposed method provides good precision and speed.

关 键 词：粒特征 贝叶斯决策 连续ADABOOST Boosting级联 人脸检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]