基于自适应推进算法的多视角机动车检测技术
|
摘要
机动车检测是模式识别、图像处理和计算机视觉领域中比较重要的研究课题,在视频监控技术、内容的图像与视频检索、机动车辆识别以及人工智能等都有着十分广泛的应用前景和实用价值。
机动车检测(即车辆检测),指对于给定的任意一幅图像,采用一定的算法和策略对其进行搜索判断其中是否存在车辆,若存在则返回车辆的位置、大小和姿态等。由于在现实生活中车辆经常以不同视角出现在视频图像中,为了实现检测方法的鲁棒性,就需要考虑车辆在各种复杂的背景中、不同方向、角度、尺度等情况下所展现出来的不同表象,即进行多视角检测。本文的主要工作如下：
(1)在特征提取与检测方面,为提高特征的计算速度,采用Harr-like特征表示图像,并引入“积分图”的概念；同时,为提高检测速度,采用自适应推进算法来选择特征组成强分类器,并采用"Cascade"策略进行检测。自适应推进算法是Viola等人提出的一种在人脸检测中应用的技术,在取得较好检测性能的同时,实现人脸的实时检测,已基本达到实时效果,可用于多视角的车辆检测及实际应用。
(2)在多视角的车辆检测器构造方面,本文采用基于Haar-Like特征和自适应推进算法相结合来构造各个视角的车辆检测器,并在训练各个视角的车辆检测器中采用Cascade方法将强分类器级联构成各个视角的最终分类器,最后在检测阶段引入视角估计进行五个不同视角的预估计,综合检测结果,得出实验结果。
(3)在增加训练样本方面,为解决训练样本不足时的情况,本文引入增加训练样本机制。通过在一幅正样本图像上应用扭曲操作,产生数张训练样本,再将此过程迭代,得到上千张训练样本,解决了训练样本不足的问题,实验结果表明增加训练样本可以提高检测效果。
本文通过采用Haar-Like特征表示图像,和采用自适应推进算法构建强分类器,并采用Cascade方法级联分类器,最后在检测阶段引入视角估计进行检测并得出检测结果。实验结果表明,采用基于Haar-Like特征和自适应推进算法能解决机动车多视角的问题,并达到检测的目的。
Vehicle detection is an important research theme in the topic of Pattern Recognition, Image processing and Computer Vision; it has a wide application prospects and practical value in many fields such as video surveillance, content image and video retrieval, automatic vehicle recognition and artificial intelligence, etc.
Given an arbitrary image, adopt certain algorithm or strategy to search in order to determine the existence of vehicles. If exist, then return the position, size, view of vehicle. As in real life with different perspectives vehicles often appear in the video image, and in order to improve the robustness of detection, we have to consider different appearance which the vehicle in a variety of complex backgrounds, different direction, angle, scale and other circumstances revealed. Namely, Multi-View Detection. In this thesis, the major work as follows:
(1) In the aspect of feature extraction and detection, in order to improve the compute speed of feature, use Harr-like feature to presentation image and introduce concept of "integral image". Besides, to improve face detection rate, use AdaBoost technique to choose features for compose strong classifier and applying "Cascade" strategy for detection. AdaBoost algorithm is a technology which viola applied in face detection. This method, obtain good detect performance and realize the face of real-time detection. It basically achieve real-time. Therefore, it can be used in Multi-View Vehicle detection and practical application.
(2) In the aspect of construct various perspectives of vehicle detector, we adopt Haar-Like Features and AdaBoost learning algorithm to construct the various perspectives of the vehicle detector. Use cascade method to constitute a strong classifiers various perspectives of the final classifier in the training multi-view vehicle detector. Finally, in the testing phase, introduce view estimate to predict five different perspectives, then comprehensive test results and obtained experimental results.
(3) In the aspect of increase training samples, to solve the problem of insufficient training samples, we introduce a mechanism to increase the training samples. Create training samples from one image applying distortions, then generate a few training samples, and then iterate this process. Therefore, we can get thousands of training samples and solve the problem of insufficient training samples. Experimental results show that increasing the training samples can improve detection.
This thesis adopts Haar-Like Features to presentation image and use AdaBoost algorithm to construct strong classifiers. Besides, use cascade method to constitute classifiers. Finally, in the testing phase, introduce view estimate to detect and obtain experimental results. The results shows that adopt Haar-Like Features and AdaBoost learning algorithm can solve the problem of vehicle multi-view and achieve the destination of detection.
机动车检测(即车辆检测),指对于给定的任意一幅图像,采用一定的算法和策略对其进行搜索判断其中是否存在车辆,若存在则返回车辆的位置、大小和姿态等。由于在现实生活中车辆经常以不同视角出现在视频图像中,为了实现检测方法的鲁棒性,就需要考虑车辆在各种复杂的背景中、不同方向、角度、尺度等情况下所展现出来的不同表象,即进行多视角检测。本文的主要工作如下：
(1)在特征提取与检测方面,为提高特征的计算速度,采用Harr-like特征表示图像,并引入“积分图”的概念；同时,为提高检测速度,采用自适应推进算法来选择特征组成强分类器,并采用"Cascade"策略进行检测。自适应推进算法是Viola等人提出的一种在人脸检测中应用的技术,在取得较好检测性能的同时,实现人脸的实时检测,已基本达到实时效果,可用于多视角的车辆检测及实际应用。
(2)在多视角的车辆检测器构造方面,本文采用基于Haar-Like特征和自适应推进算法相结合来构造各个视角的车辆检测器,并在训练各个视角的车辆检测器中采用Cascade方法将强分类器级联构成各个视角的最终分类器,最后在检测阶段引入视角估计进行五个不同视角的预估计,综合检测结果,得出实验结果。
(3)在增加训练样本方面,为解决训练样本不足时的情况,本文引入增加训练样本机制。通过在一幅正样本图像上应用扭曲操作,产生数张训练样本,再将此过程迭代,得到上千张训练样本,解决了训练样本不足的问题,实验结果表明增加训练样本可以提高检测效果。
本文通过采用Haar-Like特征表示图像,和采用自适应推进算法构建强分类器,并采用Cascade方法级联分类器,最后在检测阶段引入视角估计进行检测并得出检测结果。实验结果表明,采用基于Haar-Like特征和自适应推进算法能解决机动车多视角的问题,并达到检测的目的。
Vehicle detection is an important research theme in the topic of Pattern Recognition, Image processing and Computer Vision; it has a wide application prospects and practical value in many fields such as video surveillance, content image and video retrieval, automatic vehicle recognition and artificial intelligence, etc.
Given an arbitrary image, adopt certain algorithm or strategy to search in order to determine the existence of vehicles. If exist, then return the position, size, view of vehicle. As in real life with different perspectives vehicles often appear in the video image, and in order to improve the robustness of detection, we have to consider different appearance which the vehicle in a variety of complex backgrounds, different direction, angle, scale and other circumstances revealed. Namely, Multi-View Detection. In this thesis, the major work as follows:
(1) In the aspect of feature extraction and detection, in order to improve the compute speed of feature, use Harr-like feature to presentation image and introduce concept of "integral image". Besides, to improve face detection rate, use AdaBoost technique to choose features for compose strong classifier and applying "Cascade" strategy for detection. AdaBoost algorithm is a technology which viola applied in face detection. This method, obtain good detect performance and realize the face of real-time detection. It basically achieve real-time. Therefore, it can be used in Multi-View Vehicle detection and practical application.
(2) In the aspect of construct various perspectives of vehicle detector, we adopt Haar-Like Features and AdaBoost learning algorithm to construct the various perspectives of the vehicle detector. Use cascade method to constitute a strong classifiers various perspectives of the final classifier in the training multi-view vehicle detector. Finally, in the testing phase, introduce view estimate to predict five different perspectives, then comprehensive test results and obtained experimental results.
(3) In the aspect of increase training samples, to solve the problem of insufficient training samples, we introduce a mechanism to increase the training samples. Create training samples from one image applying distortions, then generate a few training samples, and then iterate this process. Therefore, we can get thousands of training samples and solve the problem of insufficient training samples. Experimental results show that increasing the training samples can improve detection.
This thesis adopts Haar-Like Features to presentation image and use AdaBoost algorithm to construct strong classifiers. Besides, use cascade method to constitute classifiers. Finally, in the testing phase, introduce view estimate to detect and obtain experimental results. The results shows that adopt Haar-Like Features and AdaBoost learning algorithm can solve the problem of vehicle multi-view and achieve the destination of detection.
引文
[1]高美国.智能运输系统中的雷达车辆检测器[J].北京理工大学学报,2001,2(4)：5-9
[2]朱海涛.一种基于地感线圈传感器的车检器设计[J].电子工程,2002,3(5)：39-42
[3]吴祖杨,沈宏庆,都思丹.背景提取基础上运动车辆视频检测[J].交通与计算机,2003,21(6)：18-20
[4]P.Viola, M.Jones. Rapid Object Detection using a Boosted Cascade of Simple Features[J]. In proceedings IEEE conf. on Computer Vision and Pattern Recognition, Kauai, Hawaii, USA,2001:511-518
[5]Li SZ,Zhu L, Zhang ZQ,Zhang HJ. Learning to Detect Multi-View Faces in Real-Time[C]. In:Proceedings of the 2nd International Conference on Development and Learning, Washington DC.June,2002
[6]Li SZ,ZhU L,Zhang ZQ,Blake A, etal. Statistical Learning of Multi-View Face Detection[C]. In:Proceedings of the 7th European Conference on Computer Vision. Copenhagen,Denmark,May,2002
[7]Collins R, Lipton A J, Kanade T, etal. A system for video surveillance and monitoring:VSAM final report[J]. Technical Report:CMU-RI-TR-00-12,Carnegie Melon University, Pittsburgh,Peen,America,2000
[8]Collins R,Lipton A,Kanade T. Introduction to the special section on video surveillance[J]. IEEE Trans.Pattern Analysis and Machine Intelligence,2000: 22(8):745-746
[9]http://www.ci.nyc.ny.us/html/nypd/html/transportation/traffic.html
[10]http://www.jhsafe.com/safe/article
[11]Matsuyama T. Cooperative Distributed Vision[J]. In Proceedings of DARPA Image Understanding Workshop,1998:365-384
[12]Milos Stojmenovic. Real time object detection in images based on an AdaBoost machine learning approach and a small training set[D]. Ottawa-Carleton Institute for Computer Science,2005(6)
[13]Chang Huang, Haizhou Ai, Bo Wu, etal. Boosting Nested Cascade Detector for Multi-View Face Detection[J]. IEEE Computer Society,2004:415-418
[14]J.Thureson, S.Carlsson, Finding Object Categories in Cluttered Image Using Minimal Shape Prototypes[C],13th Scandinavian Conference on Image Analysis SCIA, Goteborg, Sweden, pp.2003:1122-1129
[15]V.S.Petrovic, T.F.Cootes, Analysis of Features for Rigid Structure Vehicle Type Recognition[J], Proc. British Machine Vision Conference, Vol.2,pp.2004:587-596
[16]E.Osuna, Applying SVMs to face detection[J], IEEE Intelligent Systems, July/August, 1998:23-26
[17]D.Le, S.Satoh, Fusion of Local and Global Features for Efficient Object Detection[J], IS & T/SPIE Symposium on Electronic Imaging,2005
[18]Bartlett, M.S., Littlewort,G., Fasel,I., Movellan,J.R., Real time face detection and expression recognition:Devepopment and application to human-computer interaction[C],CVPR Workshop on Computer Vision and Pattern Recognition for Human-Computer Interaction, at IEEE CVPR, Madison, Wisconsin, June 17,2003
[19]P.Viloa, M.Jones. Robust Real-Time Face Detection[J]. International Journal of Computer Vision.2004,57(2):137-154
[20]Y.Freund, R.E.Schapire, A decision-theoretic generalization on on-line learning and an application to boosting[C], Proc.2nd European Conference on Computational Learning Theory(Eurocolt95),Barcelona, Spain,23-37,1995; Journal of Computer and System Sciences,55(1):pp.119-139, August 1997
[21]Y.Freund, R.E.Schapire, A Short Introduction to Boosting, Journal of Japanese Society for Artificial Intelligence[J],14(5):pp.771-780, September,1999
[22]R.Schapire, Y. Singer, Improved Boosting Algorithms Using Confidence-rated Predictions[J], Machine Learning,37(3):pp.297-336, December 1999
[23]P.Viola, M.Jones. Robust Real-time Object Detection[J]. Cambridge Research Laboratory, Technical Report Series. CRL 2001/01
[24]R.Lienhart, J.Maydt, An extended set of Haar-Like Features for Rapid Object Detection[J],Proc.IEEE Int.Conf.Image Processing,vol.1,pp.2002:900-903
[25]Y.Freund,R.E.Schapire, A decision-theoretic generalization on on-line learning and an application to boosting[C], Proc.2nd European Conference on Computational Learning Theory(Eurocolt95), Barcelona, Spain,23-37,1995; Journal of Computer and System Sciences,55(1).119-139,August 1997
[26]K.Levi, Y.Weiss, Learning Object Detection from a Small Number of Examples:the Improtance of Good Features[C], International Conference on Computer Vision and Pattern Recognition(CVPR),Vol.2,pp.2004:53-60
[27]H.Luo, J.Yen, D.Tretter, An Efficient Automatic Redeye Detection and Correction Algorithm[C],17th IEEE International Conference on Pattern Recognition, (ICPR'04), Cambridge UK Volume 2, August 23-26, pp.2004:883-886
[28]N.Bredeche, J.D.Zucker.WM-plic:combiner des classeurs issus derepresentation differentes pour une tache didentification dobjets par un robot situe[C]. Proceedings of La Conference dApprentissage 2003(Cap), pp.17-29. Laval,France,2003
[29]M.A.Maloof, P.Langley, T.O.Binford, etal, Improved rooftop detection in aerial images with machine learning[J], Machine Learning 53:pp.2003:157-191
[30]Schapire R E. The strength of weak learnability[J]. Machine Learning, 1990,5(2):197-227
[31]Freund Y. Boosting a Weak Learning Algorithm by Majority[J]. Information and Computation,1995,141(2):256-285
[32]M Turk, A Pentland. Eigenfaces for Recognition[J]. J.cognitive Neuroscienee,1991, 3(1):71-86
[33]Freund Y, Schapire R E. A Decision-theoretic Generalization of Online Learning and an Application to Boosting[J]. Journal of Computer and System Sciences, 1997,55(1):119-139
[34]Valiant L G. A Theory of the Learnable[J]. Communications of the ACM, 1984,27(11):1134-1142
[35]Michael Kearns, Leslie G. Valiant. Learning Boolean formulae or finite automata is as hard as factoring[J]. Technical Report TR-14-88, Harvard University Aiken Computation Laboratory, August 1988
[36]Michael Kearns, Leslie G. Valiant. Cryptographic limitations on learning Boolean formulae and finite automata[J]. Journal of the Association for Computing Machinery, 41(1):67-95, January 1994
[37]L.G. Valiant. A theory of the learnable[J]. Communications of the ACM, 27(11):1134-1142, November 1984
[38]Kearns M, Valiant L G. Learning Boolean Formulae or Factoring[R]. Technical Report TR-1488, Cambridge, MA:Havard University Aiken Computation Laboratory,1998
[39]Kearns M, Valiant L G. Crytographic Limitation on Learning Boolean Formulae and Finite Automata[C]. In:Proceedings of the 21st Annual ACM Symposium on Theory of Computing, New York, NY:ACM press,1989:433-444
[40]Yoav Freund, Robert E. Schapire. Experiments with a New Boosting Algorithm[C]. In:International Conference on Machine Learning,1996:148-156
[41]李斌.分类器优化算法的研究[D].复旦大学,2003
[42]Lienhart R, Kuranov A, V Pisarevsky. Empirical analysis of detection cascades of boosted classifiers for rapid object detection[C]. DAGM'03 25th Pattern Recognition Symposium 2003
[43]黄向生.基于Boosting学习的自动人脸识别算法研究[D].中国科学院自动化研究所,2005
[44]Duy-Dinh Le, Shin'ichi Sotoh. Ent-Boost:Boosting using entropy measures for robust object detection[J]. Pattern Recognition Letters,2007,28:1083-1090
[45]严云洋,郭志波,杨静宇.基于双阈值的增强型AdaBoost决速算法[J].计算机工程,2007,33(21)：172-174
[46]Schapire R E, Singer Y. Improved boosting algorithms using confidence-rated predictions[C]. In:Proceedings of 11th Annual Conference on Computational Learning Theory,1998:80-91
[47]Schapire R E, Freund Y, Bartlett P, etal. Boosting the Margin:a New Explanation for the Effectiveness of Voting Methods[J]. The Annals of Statistics,1998,26(5):1651-1686
[48]Honik K. Approximation capabilities of multilayer feedforward networks[J]. Neural Networks,1991,4:551-557
[49]阎平凡,黄端旭.人工神经-模型[J].分析与应用.合肥：安徽教育出版社,1993
[50]Freund Y, Schapire R E, A Decision-Theoretic Generalization of Online Learning and Application to Boosting[J]. Journal of Computer and System Scieneces,1997, 55(1):119-139
[51]C.Nakajima, M.Pontil, B.Heisele, etal. People Recognition in Image Sequences by Supervised Learning[J].A.I.Memo No.1688,June,2000
[52]C.Papageorgiou, M.Oren, T.Poggio. A general framework for Object Detection[C]. In International Conference on Computer Vision,1998
[53]王英明.基于AdaBoost算法的人脸检测研究[D].江苏.南京理工大学.2008
[54]杜杰.基于AdaBoost的快速人脸检测算法若干问题研究[D].江苏.南京理工大学.2007
[2]朱海涛.一种基于地感线圈传感器的车检器设计[J].电子工程,2002,3(5)：39-42
[3]吴祖杨,沈宏庆,都思丹.背景提取基础上运动车辆视频检测[J].交通与计算机,2003,21(6)：18-20
[4]P.Viola, M.Jones. Rapid Object Detection using a Boosted Cascade of Simple Features[J]. In proceedings IEEE conf. on Computer Vision and Pattern Recognition, Kauai, Hawaii, USA,2001:511-518
[5]Li SZ,Zhu L, Zhang ZQ,Zhang HJ. Learning to Detect Multi-View Faces in Real-Time[C]. In:Proceedings of the 2nd International Conference on Development and Learning, Washington DC.June,2002
[6]Li SZ,ZhU L,Zhang ZQ,Blake A, etal. Statistical Learning of Multi-View Face Detection[C]. In:Proceedings of the 7th European Conference on Computer Vision. Copenhagen,Denmark,May,2002
[7]Collins R, Lipton A J, Kanade T, etal. A system for video surveillance and monitoring:VSAM final report[J]. Technical Report:CMU-RI-TR-00-12,Carnegie Melon University, Pittsburgh,Peen,America,2000
[8]Collins R,Lipton A,Kanade T. Introduction to the special section on video surveillance[J]. IEEE Trans.Pattern Analysis and Machine Intelligence,2000: 22(8):745-746
[9]http://www.ci.nyc.ny.us/html/nypd/html/transportation/traffic.html
[10]http://www.jhsafe.com/safe/article
[11]Matsuyama T. Cooperative Distributed Vision[J]. In Proceedings of DARPA Image Understanding Workshop,1998:365-384
[12]Milos Stojmenovic. Real time object detection in images based on an AdaBoost machine learning approach and a small training set[D]. Ottawa-Carleton Institute for Computer Science,2005(6)
[13]Chang Huang, Haizhou Ai, Bo Wu, etal. Boosting Nested Cascade Detector for Multi-View Face Detection[J]. IEEE Computer Society,2004:415-418
[14]J.Thureson, S.Carlsson, Finding Object Categories in Cluttered Image Using Minimal Shape Prototypes[C],13th Scandinavian Conference on Image Analysis SCIA, Goteborg, Sweden, pp.2003:1122-1129
[15]V.S.Petrovic, T.F.Cootes, Analysis of Features for Rigid Structure Vehicle Type Recognition[J], Proc. British Machine Vision Conference, Vol.2,pp.2004:587-596
[16]E.Osuna, Applying SVMs to face detection[J], IEEE Intelligent Systems, July/August, 1998:23-26
[17]D.Le, S.Satoh, Fusion of Local and Global Features for Efficient Object Detection[J], IS & T/SPIE Symposium on Electronic Imaging,2005
[18]Bartlett, M.S., Littlewort,G., Fasel,I., Movellan,J.R., Real time face detection and expression recognition:Devepopment and application to human-computer interaction[C],CVPR Workshop on Computer Vision and Pattern Recognition for Human-Computer Interaction, at IEEE CVPR, Madison, Wisconsin, June 17,2003
[19]P.Viloa, M.Jones. Robust Real-Time Face Detection[J]. International Journal of Computer Vision.2004,57(2):137-154
[20]Y.Freund, R.E.Schapire, A decision-theoretic generalization on on-line learning and an application to boosting[C], Proc.2nd European Conference on Computational Learning Theory(Eurocolt95),Barcelona, Spain,23-37,1995; Journal of Computer and System Sciences,55(1):pp.119-139, August 1997
[21]Y.Freund, R.E.Schapire, A Short Introduction to Boosting, Journal of Japanese Society for Artificial Intelligence[J],14(5):pp.771-780, September,1999
[22]R.Schapire, Y. Singer, Improved Boosting Algorithms Using Confidence-rated Predictions[J], Machine Learning,37(3):pp.297-336, December 1999
[23]P.Viola, M.Jones. Robust Real-time Object Detection[J]. Cambridge Research Laboratory, Technical Report Series. CRL 2001/01
[24]R.Lienhart, J.Maydt, An extended set of Haar-Like Features for Rapid Object Detection[J],Proc.IEEE Int.Conf.Image Processing,vol.1,pp.2002:900-903
[25]Y.Freund,R.E.Schapire, A decision-theoretic generalization on on-line learning and an application to boosting[C], Proc.2nd European Conference on Computational Learning Theory(Eurocolt95), Barcelona, Spain,23-37,1995; Journal of Computer and System Sciences,55(1).119-139,August 1997
[26]K.Levi, Y.Weiss, Learning Object Detection from a Small Number of Examples:the Improtance of Good Features[C], International Conference on Computer Vision and Pattern Recognition(CVPR),Vol.2,pp.2004:53-60
[27]H.Luo, J.Yen, D.Tretter, An Efficient Automatic Redeye Detection and Correction Algorithm[C],17th IEEE International Conference on Pattern Recognition, (ICPR'04), Cambridge UK Volume 2, August 23-26, pp.2004:883-886
[28]N.Bredeche, J.D.Zucker.WM-plic:combiner des classeurs issus derepresentation differentes pour une tache didentification dobjets par un robot situe[C]. Proceedings of La Conference dApprentissage 2003(Cap), pp.17-29. Laval,France,2003
[29]M.A.Maloof, P.Langley, T.O.Binford, etal, Improved rooftop detection in aerial images with machine learning[J], Machine Learning 53:pp.2003:157-191
[30]Schapire R E. The strength of weak learnability[J]. Machine Learning, 1990,5(2):197-227
[31]Freund Y. Boosting a Weak Learning Algorithm by Majority[J]. Information and Computation,1995,141(2):256-285
[32]M Turk, A Pentland. Eigenfaces for Recognition[J]. J.cognitive Neuroscienee,1991, 3(1):71-86
[33]Freund Y, Schapire R E. A Decision-theoretic Generalization of Online Learning and an Application to Boosting[J]. Journal of Computer and System Sciences, 1997,55(1):119-139
[34]Valiant L G. A Theory of the Learnable[J]. Communications of the ACM, 1984,27(11):1134-1142
[35]Michael Kearns, Leslie G. Valiant. Learning Boolean formulae or finite automata is as hard as factoring[J]. Technical Report TR-14-88, Harvard University Aiken Computation Laboratory, August 1988
[36]Michael Kearns, Leslie G. Valiant. Cryptographic limitations on learning Boolean formulae and finite automata[J]. Journal of the Association for Computing Machinery, 41(1):67-95, January 1994
[37]L.G. Valiant. A theory of the learnable[J]. Communications of the ACM, 27(11):1134-1142, November 1984
[38]Kearns M, Valiant L G. Learning Boolean Formulae or Factoring[R]. Technical Report TR-1488, Cambridge, MA:Havard University Aiken Computation Laboratory,1998
[39]Kearns M, Valiant L G. Crytographic Limitation on Learning Boolean Formulae and Finite Automata[C]. In:Proceedings of the 21st Annual ACM Symposium on Theory of Computing, New York, NY:ACM press,1989:433-444
[40]Yoav Freund, Robert E. Schapire. Experiments with a New Boosting Algorithm[C]. In:International Conference on Machine Learning,1996:148-156
[41]李斌.分类器优化算法的研究[D].复旦大学,2003
[42]Lienhart R, Kuranov A, V Pisarevsky. Empirical analysis of detection cascades of boosted classifiers for rapid object detection[C]. DAGM'03 25th Pattern Recognition Symposium 2003
[43]黄向生.基于Boosting学习的自动人脸识别算法研究[D].中国科学院自动化研究所,2005
[44]Duy-Dinh Le, Shin'ichi Sotoh. Ent-Boost:Boosting using entropy measures for robust object detection[J]. Pattern Recognition Letters,2007,28:1083-1090
[45]严云洋,郭志波,杨静宇.基于双阈值的增强型AdaBoost决速算法[J].计算机工程,2007,33(21)：172-174
[46]Schapire R E, Singer Y. Improved boosting algorithms using confidence-rated predictions[C]. In:Proceedings of 11th Annual Conference on Computational Learning Theory,1998:80-91
[47]Schapire R E, Freund Y, Bartlett P, etal. Boosting the Margin:a New Explanation for the Effectiveness of Voting Methods[J]. The Annals of Statistics,1998,26(5):1651-1686
[48]Honik K. Approximation capabilities of multilayer feedforward networks[J]. Neural Networks,1991,4:551-557
[49]阎平凡,黄端旭.人工神经-模型[J].分析与应用.合肥：安徽教育出版社,1993
[50]Freund Y, Schapire R E, A Decision-Theoretic Generalization of Online Learning and Application to Boosting[J]. Journal of Computer and System Scieneces,1997, 55(1):119-139
[51]C.Nakajima, M.Pontil, B.Heisele, etal. People Recognition in Image Sequences by Supervised Learning[J].A.I.Memo No.1688,June,2000
[52]C.Papageorgiou, M.Oren, T.Poggio. A general framework for Object Detection[C]. In International Conference on Computer Vision,1998
[53]王英明.基于AdaBoost算法的人脸检测研究[D].江苏.南京理工大学.2008
[54]杜杰.基于AdaBoost的快速人脸检测算法若干问题研究[D].江苏.南京理工大学.2007