复杂交通场景中基于视频的行人检测与跟踪若干关键问题研究
|
摘要
基于视频的行人检测、跟踪与计数研究是计算机视觉领域的一个热点研究方向，该研究在交通安全、实时客流统计、智能监控等领域都有着广泛的应用前景。然而，由于检测环境中光照条件的差异性，行人与非行人样本在数量上的非均衡问题，以及遮挡和干扰等其他因素的影响，该研究仍面临着许多困难，检测精度及检测方法的稳定性都有待进一步提高。
本论文研究行人检测与跟踪中的若干关键问题，主要包括：有效的行人特征提取算法；同时考虑样本非均衡性和弱分类器多样性的分类算法；分类算法中关键参数取值动态寻优方法；高效的分类器搭建结构实现；基于颜色、纹理及空间信息的快速行人跟踪方法；多行人跟踪数据关联方法以及基于行人检测与多行人跟踪的行人计数方法。具体工作如下：
1.提出一种双层差异特征提取算法。该特征的上层子特征由改进的ABH行人特征构成，ABH特征通过引入局部二元模式的思想对二值化后的Haar特征加以组合，在保持特征实时性高的同时，有效的增强特征的光照不变性，用于对行人目标进行快速检测定位；然后，利用Edgelet特征作为下层子特征验证ABH特征所提取行人的有效性，降低虚警和遮挡对检测的影响。
2.提出一种基于CS-SVM的非均衡Gentle AdaBoost分类算法（DGACS分类算法）。该算法首先引入代价敏感的SVM (CS-SVM)分类算法搭建弱分类器，并利用非均衡Gentle Adaboost算法规则将弱分类器组合为强分类器，使其有利于解决正负样本数量不均等的行人分类问题；然后，通过动态调整CS-SVM的核函数参数取值来实现弱分类器多样化的目的，并制定相应规则剔除相似度高的弱分类器，增强弱分类器之间的差异性，从而进一步提高算法的分类性能。
3.由于代价敏感值的选取对DGACS分类算法的性能影响较大，提出一种动态代价敏感参数寻优方法来选取适宜的代价敏感值。该方法利用本文提出的一种新的优化算法——基于T变异的混沌粒子群算法，以正负样本正确分类的最佳折中作为寻优原则，在代价敏感权重值的取值区域内动态寻找全局最优解。
4.建立一种基于分级树状结构的组合分类器搭建方式，采用“由粗到精”的检测思想来逐级检测行人。其中，粗级组合分类器采用完全二叉树结构组合分类器，用来迅速排除大量简单易分类的负样本，筛选出正样本和疑似正样本的负样本作为候选对象；精级组合分类器采用一个串联树结构，用来对筛选后的候选对象进一步的精确分类。
5.提出一种基于融合颜色、纹理及对应子空间信息的快速粒子滤波行人跟踪算法。首先，提取目标行人的空间信息并细化成头部、上身、腿部三个局部子区域；其次，利用改进的纹理及颜色信息提取算法提取对应目标子区域中的联合纹理、颜色信息；最后，通过基于空间划分的颜色纹理相似度指标来判断跟踪目标的位置，实现准确跟踪。考虑到多线索信息融合算法所需粒子数目较大、计算效率低的问题，提出使用波面扫描的积分直方图计算方法来提高运算速度。
6.对基于k-best MHT算法的多行人数据关联算法进行改进，通过改进目标航迹的生成及选择方式，从源头上排除部分可信度低的航迹；对假设管理环节加以改进，在假设矩阵生成前先对量测进行分组，并以组为单位进行假设矩阵的修剪与合并，降低所产生的假设矩阵的维数。在行人检测及多行人跟踪的基础上搭建行人计数系统，通过检测计算图像初始帧监控区域内行人数量以及后续区域内的行人进出数量计算得出监控区域内实时行人数量，实现区域客流量信息的实时统计。
Pedestrian detection，tracking and count based on video is a research focus in computervision, and it has a wide range of applications in traffic safety, real-time traffic statistics,intelligent surveillance, etc. However, because of the different light conditions of thebackground environment, the disequilibrium of the samples in scene, occlusion andinterference, pedestrian detection，tracking and count is still facing many difficulties, whichaccuracy and stability of detection method should be further improved.
Some key issues of pedestrian detection and tracking are researched in this dissertation,which mainly include: the algorithm of extracts high robustness pedestrian feature, theclassification algorithm consider the problem of imbalanced samples and the diversity ofweak classifiers at the same time and the dynamic optimization method for the key parameters,high efficiency of the classifier structure, fast pedestrian tracking based on information ofcolor, texture and space, data correlation of multi pedestrian tracking and pedestrian countmethod based on pedestrian detection and multi pedestrian tracking. Specific studies asfollows:
1. A bilayer difference feature extraction algorithm is proposed. The upper layer ofbilayer difference feature is constituted by improved Assembled Binary Haar (ABH) feature,which is inspired by the idea of Local Binary Pattern (LBP) and combine binary Haar featuresthrough the LBP rule. ABH feature could enhance the ability of illumination invariance andkeep high real-time performance, which is used for fast pedestrian detection and location;then, Edgelet feature is used for the lower layer of the bilayer difference feature and check thevalidity of pedestrian which is detected by ABH feature, reduce the influence of false alarmand occlusion to detection.
2. A Disequilibrium Gentle Adaboost with classifiers algorithm is proposed. In order tosolve the problem of class-imbalanced in pedestrian detection, this new algorithm lead inCost-Sensitive SVM (CS-SVM) as weak classifier and use Disequilibrium Gentle AdaBoostalgorithm to compose weak classifiers into strong classifiers; in addition, through dynamicadjust the values of the kernel parameter in CS-SVM and get a series of component classifiersdifferent with each other, and measure the diversity of component classifiers, then discardpoor and similar classifiers, which could improve diversity among classifiers and theclassification performance of the algorithm.
3. Because the parameters of cost-sensitive is very important for the classificationperformance of DGACS, a dynamic optimization method about the parameters ofcost-sensitive is proposed for choosing the appropriate parameters. This method uses a newoptimization algorithm which is proposed in this dissertation (Chaotic Particle SwarmOptimization with T mutation). The best compromise about the correct classification of thepositive and negative samples is used for optimization principle, and dynamic choose theglobal optimal solution in the range of the cost-sensitive performance.
4. A combination classifiers based on hierarchical tree structure is established, whichadopts the ideology of “coarse-to-fine” to detect pedestrian step by step. Among them, thecoarse combinations classifier employs a full binary tree structure. This structure is used forrejecting numerous obvious negative samples, and selecting positive samples and somehidden negative samples as candidates; the fine combinations classifier utilize a series treestructure, which used to classify more precise from candidates.
5. A fast particle filter pedestrian tracking method based on color, texture andcorresponding space information. Firstly, we extract space information of object pedestrianand disintegrate it into three local regions (head, upper body and leg); next, employ theimproved texture and color information extract algorithm to get the joint texture and colorinformation from the corresponding sub-region; finally, determine the position of object bycolor-texture similarity indicator based on space division, and get the result of accuratelytrack. In consideration of the multi thread information fusion algorithm need a larger numberof particles, and reduce the computational efficiency. Therefore, a wave integral histogramalgorithm is proposed for improving arithmetic speed.
6. A multi-pedestrian tracking algorithm based on k-best MHT is improved. Firstly, forexcluding part of the tracks which have low credibility from the source, we improve the wayof generate and select the objective tracks; secondly, the link of hypothesis management ismend. In order to reduce the dimensions of hypothesis matrix, measurements are groupedbefore hypothesis matrix generation, and use group as a unit to trim or merge matrixgenerations. A pedestrian counting system is set up which is based on pedestrian detection andmulti-pedestrian tracking. The system obtains the number of real-time pedestrians inmonitoring area by detect the number of pedestrians who are in the initial frame and calculatethe number of pedestrians who pass in and out monitoring area in the follow-up frames, which could get the real-time passenger flow information statistics in the detection regions.
本论文研究行人检测与跟踪中的若干关键问题，主要包括：有效的行人特征提取算法；同时考虑样本非均衡性和弱分类器多样性的分类算法；分类算法中关键参数取值动态寻优方法；高效的分类器搭建结构实现；基于颜色、纹理及空间信息的快速行人跟踪方法；多行人跟踪数据关联方法以及基于行人检测与多行人跟踪的行人计数方法。具体工作如下：
1.提出一种双层差异特征提取算法。该特征的上层子特征由改进的ABH行人特征构成，ABH特征通过引入局部二元模式的思想对二值化后的Haar特征加以组合，在保持特征实时性高的同时，有效的增强特征的光照不变性，用于对行人目标进行快速检测定位；然后，利用Edgelet特征作为下层子特征验证ABH特征所提取行人的有效性，降低虚警和遮挡对检测的影响。
2.提出一种基于CS-SVM的非均衡Gentle AdaBoost分类算法（DGACS分类算法）。该算法首先引入代价敏感的SVM (CS-SVM)分类算法搭建弱分类器，并利用非均衡Gentle Adaboost算法规则将弱分类器组合为强分类器，使其有利于解决正负样本数量不均等的行人分类问题；然后，通过动态调整CS-SVM的核函数参数取值来实现弱分类器多样化的目的，并制定相应规则剔除相似度高的弱分类器，增强弱分类器之间的差异性，从而进一步提高算法的分类性能。
3.由于代价敏感值的选取对DGACS分类算法的性能影响较大，提出一种动态代价敏感参数寻优方法来选取适宜的代价敏感值。该方法利用本文提出的一种新的优化算法——基于T变异的混沌粒子群算法，以正负样本正确分类的最佳折中作为寻优原则，在代价敏感权重值的取值区域内动态寻找全局最优解。
4.建立一种基于分级树状结构的组合分类器搭建方式，采用“由粗到精”的检测思想来逐级检测行人。其中，粗级组合分类器采用完全二叉树结构组合分类器，用来迅速排除大量简单易分类的负样本，筛选出正样本和疑似正样本的负样本作为候选对象；精级组合分类器采用一个串联树结构，用来对筛选后的候选对象进一步的精确分类。
5.提出一种基于融合颜色、纹理及对应子空间信息的快速粒子滤波行人跟踪算法。首先，提取目标行人的空间信息并细化成头部、上身、腿部三个局部子区域；其次，利用改进的纹理及颜色信息提取算法提取对应目标子区域中的联合纹理、颜色信息；最后，通过基于空间划分的颜色纹理相似度指标来判断跟踪目标的位置，实现准确跟踪。考虑到多线索信息融合算法所需粒子数目较大、计算效率低的问题，提出使用波面扫描的积分直方图计算方法来提高运算速度。
6.对基于k-best MHT算法的多行人数据关联算法进行改进，通过改进目标航迹的生成及选择方式，从源头上排除部分可信度低的航迹；对假设管理环节加以改进，在假设矩阵生成前先对量测进行分组，并以组为单位进行假设矩阵的修剪与合并，降低所产生的假设矩阵的维数。在行人检测及多行人跟踪的基础上搭建行人计数系统，通过检测计算图像初始帧监控区域内行人数量以及后续区域内的行人进出数量计算得出监控区域内实时行人数量，实现区域客流量信息的实时统计。
Pedestrian detection，tracking and count based on video is a research focus in computervision, and it has a wide range of applications in traffic safety, real-time traffic statistics,intelligent surveillance, etc. However, because of the different light conditions of thebackground environment, the disequilibrium of the samples in scene, occlusion andinterference, pedestrian detection，tracking and count is still facing many difficulties, whichaccuracy and stability of detection method should be further improved.
Some key issues of pedestrian detection and tracking are researched in this dissertation,which mainly include: the algorithm of extracts high robustness pedestrian feature, theclassification algorithm consider the problem of imbalanced samples and the diversity ofweak classifiers at the same time and the dynamic optimization method for the key parameters,high efficiency of the classifier structure, fast pedestrian tracking based on information ofcolor, texture and space, data correlation of multi pedestrian tracking and pedestrian countmethod based on pedestrian detection and multi pedestrian tracking. Specific studies asfollows:
1. A bilayer difference feature extraction algorithm is proposed. The upper layer ofbilayer difference feature is constituted by improved Assembled Binary Haar (ABH) feature,which is inspired by the idea of Local Binary Pattern (LBP) and combine binary Haar featuresthrough the LBP rule. ABH feature could enhance the ability of illumination invariance andkeep high real-time performance, which is used for fast pedestrian detection and location;then, Edgelet feature is used for the lower layer of the bilayer difference feature and check thevalidity of pedestrian which is detected by ABH feature, reduce the influence of false alarmand occlusion to detection.
2. A Disequilibrium Gentle Adaboost with classifiers algorithm is proposed. In order tosolve the problem of class-imbalanced in pedestrian detection, this new algorithm lead inCost-Sensitive SVM (CS-SVM) as weak classifier and use Disequilibrium Gentle AdaBoostalgorithm to compose weak classifiers into strong classifiers; in addition, through dynamicadjust the values of the kernel parameter in CS-SVM and get a series of component classifiersdifferent with each other, and measure the diversity of component classifiers, then discardpoor and similar classifiers, which could improve diversity among classifiers and theclassification performance of the algorithm.
3. Because the parameters of cost-sensitive is very important for the classificationperformance of DGACS, a dynamic optimization method about the parameters ofcost-sensitive is proposed for choosing the appropriate parameters. This method uses a newoptimization algorithm which is proposed in this dissertation (Chaotic Particle SwarmOptimization with T mutation). The best compromise about the correct classification of thepositive and negative samples is used for optimization principle, and dynamic choose theglobal optimal solution in the range of the cost-sensitive performance.
4. A combination classifiers based on hierarchical tree structure is established, whichadopts the ideology of “coarse-to-fine” to detect pedestrian step by step. Among them, thecoarse combinations classifier employs a full binary tree structure. This structure is used forrejecting numerous obvious negative samples, and selecting positive samples and somehidden negative samples as candidates; the fine combinations classifier utilize a series treestructure, which used to classify more precise from candidates.
5. A fast particle filter pedestrian tracking method based on color, texture andcorresponding space information. Firstly, we extract space information of object pedestrianand disintegrate it into three local regions (head, upper body and leg); next, employ theimproved texture and color information extract algorithm to get the joint texture and colorinformation from the corresponding sub-region; finally, determine the position of object bycolor-texture similarity indicator based on space division, and get the result of accuratelytrack. In consideration of the multi thread information fusion algorithm need a larger numberof particles, and reduce the computational efficiency. Therefore, a wave integral histogramalgorithm is proposed for improving arithmetic speed.
6. A multi-pedestrian tracking algorithm based on k-best MHT is improved. Firstly, forexcluding part of the tracks which have low credibility from the source, we improve the wayof generate and select the objective tracks; secondly, the link of hypothesis management ismend. In order to reduce the dimensions of hypothesis matrix, measurements are groupedbefore hypothesis matrix generation, and use group as a unit to trim or merge matrixgenerations. A pedestrian counting system is set up which is based on pedestrian detection andmulti-pedestrian tracking. The system obtains the number of real-time pedestrians inmonitoring area by detect the number of pedestrians who are in the initial frame and calculatethe number of pedestrians who pass in and out monitoring area in the follow-up frames, which could get the real-time passenger flow information statistics in the detection regions.
引文
[1]邱俊．中国交通事故和交通伤成因、特点与趋势研究[D]．重庆：第三军医大学，2009
[2]邹铁方．车人碰撞事故再现及结果不确定性分析[D]．广州：中山大学，2010
[3]申杰．汽车-行人碰撞事故再现研究与应用[D]．上海：上海交通大学，2007
[4] M. Enzweiler, D. M. Gavrila. Monocular pedestrian detection: Survey and experiments[J]．Journal of IEEE Transactions on Pattern Analysis and Machine Intelligence，2009，31(12):2179–2195
[5] Lie Guo, Ping-Shu Ge, Ming-Heng Zhang, et al. Pedestrian detection for intelligenttransportation systems combining AdaBoost algorithm and support vector machine[J]．Expert Systems with Applications，2012，39(4):4274–4286
[6] Ramesh Jain, Rangachar Kasturi, Brian G. Schunck．Machine vision [M]．Bei jing：ChinaMachine Press，2003
[7] P.Viola, M.J.Jones, D.Snow. Detecting Pedestrians Using Patterns of Motion andAppearance [J]. Int.J.Comput.Vision,2005,63(2):153–161
[8]汤义．智能交通系统中基于视频的行人检测与跟踪方法的研究[D]．广州：华南理工大学，2010
[9]田广.基于视觉的行人检测和跟踪技术的研究[D].上海：上海交通大学，2007
[10] Wang Hongliang, Wang Jinqi, Ding Haifei, et al. Moving target detection based on theimproved Gaussian mixture model background difference method [C]//3rd internationalConference on Manufacturing Science and Engineering. Xiamen，China，2012:482-484
[11]朱慧颖，张利，李云廷.基于背景差分法的交通事件智能检测系统[J]．武汉理工大学学报，2011，33(2):79–83
[12]王永忠，梁彦，潘泉，程咏梅，赵春晖.基于自适应混合高斯模型的时空背景建模[J]．自动化学报，2009，35(4):371–378
[13]李娟，邵春福，杨励雅.基于混合高斯模型的行人检测方法[J]．吉林大学学报(工学版)，2011，41(1):41–45
[14] Zhang X, Sexton G. Automatic pedestrian counting using image processing techniques[J]. Electronics Letters,1995,31(11):863–865
[15] Gavrila D.M., Munder S. Multi-cue pedestrian detection and tracking from a movingvehicle [J]. International Journal of Computer Vision,2007,73(1):41-59
[16] Gavrila D.M., Giebel J., Munder S. Vision-based pedestrian detection: ThePROTECTOR system [C]//Proceedings of IEEE Intelligent Vehicles Symposium,Parma, Italy,2004,13-18
[17] Gavrila D.M., Franke U., W hler, C., G rzig, S. Real-time vision for intelligent vehicles[J]. IEEE Instrumentation and Measurement Magazine,2001,4(2):22–27
[18]许言午.面向行人检测的组合分类计算模型与应用研究[D].合肥：中国科学技术大学，2009
[19] Wu Bo, Nevatia Ram. Detection of multiple, partially occluded humans in a single imageby bayesian combination of edgelet part detectors [J] International Journal of ComputerVision,2007,75(2):247-266
[20] Michailovich Oleg, Tannenbaum Allen. A method for prediction and estimation oflarge-amplitude optical flows via extended Kalman filtering approach [J]. EngineeringComputations,2006,23(5):503-514
[21]裴巧娜.基于光流法的运动目标检测与跟踪技术[D].北京：北方工业大学，2009
[22]杨叶梅.基于改进光流法的运动目标检测[J]．计算机与数字工程，2011，39(9):108–111
[23]王春涛.基于背景差分法和光流法的视频动态目标检测与跟踪[J]．软件导报，2011，10(6):145–147
[24] Zimmer Henning, Bruhn Andrés, Weickert Joachim. Optic flow in harmony [J].International Journal of Computer Vision,2011,93(3):368-388
[25]米切尔.机器学习[M].北京：机械工业出版社，2003
[26] Paisitkriangkcrai Sakrapee, Shen Chunhua, Zhang Jian. Fast pedestrian detection using acascade of boosted covariance features [J]. IEEE Transactions on Circuits and Systemsfor Video Technology,2008,18(8):1140-1151
[27] Tuzel Oncel, Porikli Fatih, Meer Peter. Pedestrian detection via classification onRiemannian manifolds [J]. IEEE Transactions on Pattern Analysis And MachineIntelligence,2008,30(10):1713-1727
[28] Viola P., Jones M. Robust real-time face detection [J]. International Journal of ComputerVision,2004,52(2):137-154
[29] Viola P, Jones MJ, Snow D. Detecting pedestrians using patterns of motion andappearance [J]. International Journal of Computer Vision,2005,63(2):153-161
[30] Papageorgiou C, Poggio T. A trainable system for object detection [J]. InternationalJournal of Computer Vision,2000,38(1):15-33
[31] Lienhart Rainer, Maydt Jochen. An extended set of Haar-like features for rapid objectdetection [C]//Proceedings of IEEE International Conference on Image, Rochester, NY,United states,2002,900-903
[32] Xinyi Cui, Yazhou Liu, Shiguang Shan, Xilin Chen, Wen Gao.3D Haar-like features forpedestrian detection [C]//Proceedings of IEEE International Conference on Multimediaand Expo, Beijing, China,2007,1263-1266
[33] Dalal Navneet, Triggs Bill. Histograms of oriented gradients for human detection [C]//Proceedings of IEEE Computer Society Conference on Computer Vision and PatternRecognition, San Diego, CA, United states,2005,886-893
[34] Yangwei Pang, Yuan Yuan, Xuelong Li, Jing Pan. Efficient HOG human detection [J].Signal Processing,2011,91(2011):773-781
[35] Wang Xiaoyu, Han Tony X., Yan Shuicheng. An HOG-LBP Human Detector withPartial Occlusion Handling [C]//Proceedings of12th IEEE International Conference onComputer Vision, Kyoto, JAPAN,2009,32-39
[36] Lowe D M.. Distinctive image features from scale-invariant keypoints [J]. InternationalJournal of Computer Vision,2004,60(2):91-110
[37] Inan Tolga, Halici Ugur.3-D face recognition with local shape descriptors [J]. IEEETransactions on Information Forensics and Security,2012,7(2):577-587
[38]孙玉芬,刘钢.基于SIFT特征的视频内头部姿态估计[J]．武汉理工大学学报，2011，33(1):146–150
[39] Refregier A. Shapelets:I.A method for image analysis [J]. Monthly Notice of the RoyalAstronomical Society,2003,338(1):35-47
[40] Sabzmeydani P, Mori G. Detecting Pedestrians by Learning Shapelet Features [C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition,Minneapolis, MN,2007,1-8
[41] Yang Hee-Deok, Lee Seong-Whan. Multiple pedestrian detection and tracking based onweighted temporal texture features [C]//Proceedings of International Conference onPattern Recognition, British Machine Vis Assoc, Cambridge, ENGLAND,2004,248-251
[42] Zhang Yang, Liu Weiming, Deng Rufeng, Mo Chen. A new pedestrian detection methodbased on diverse AdaBoost-SVM optimization ensemble classifiers algorithm [J].Journal of Computational Information Systems,2011,7(14):5076-5085
[43]苏松志,李绍滋,陈淑媛,蔡国榕,吴云东.行人检测技术综述[J]．电子学报，2012，40(4):814–820
[44]朱文佳.基于机器学习的行人检测关键技术研究[D].上海：上海交通大学，2008
[45] Guo Lie, Ge, Ping-Shu, Zhang Ming-Heng, Li Lin-Hui, Zhao Yi-Bing. Pedestriandetection for intelligent transportation systems combining AdaBoost algorithm andsupport vector machine [J]. Expert Systems with Applications,2012,39(4):4274-4286
[46]李闯,丁晓青,吴佑寿.一种改进的AdaBoost算法—AD AdaBoost [J]．计算机学报，2007，30(1):103–109
[47]张彦峰，何佩琨.一种改进的AdaBoost算法—M-Asy AdaBoost [J]．北京理工大学学报，2011，31(1):64–68
[48]李文辉，倪洪印.一种改进的Adaboost训练算法[J]．吉林大学学报(理学版)，2011，49(3):498–504
[49] J. B′egard N. Allezard, P. Sayd. Real-Time Human Detection in Urban Scenes: LocalDescriptors and Classifiers Selection with AdaBoost-like Algorithms [C]//Proceedingsof IEEE Computer Society Conference on Computer Vision and Pattern RecognitionWorkshops, Anchorage, AK, United states,2008,1-8
[50]朱文佳,戚飞虎.基于Gentle Adaboost的行人检测[J]．中国图象图形学报，2007，12(10):1905–1908
[51] Vapnik V, Chapelle O. Bounds on error expectation for support vector machines [J].Neural Computation,2000,12(9):2013-2036
[52] Ignacio Parra Alonso, David Fernández Llorca, Miguel ángel Sotelo. Combination ofFeature Extraction Methods for SVM Pedestrian Detection [J]. IEEE Transactions onIntelligent Transportation Systems,2007,8(2):292-307
[53]田广,戚飞虎.移动摄像机环境下基于特征变换和SVM的分级行人检测算法[J]．电子学报，2008，36(5):1024–1028
[54] Lee Gyemin, Scott Clayton. Nested support vector machines [J]. IEEE Transactions onSignal Processing,2010,58(3):1648-1660
[55]刘冲,张均东,曾鸿.基于支持向量机的无穷维AdaBoost算法及其应用[J]．仪器仪表学报，2010，31(4):764-769
[56] Strack, Robert, Kecman Vojislav, Strack Beata, Li Qi. Sphere Support Vector Machinesfor large classification tasks [J]. Neurocomputing,2013,101(3):59-67
[57] Xu Y.W., Cao X.B., Qiao H. An Efficient Tree Classifier Ensemble-based Approach forPedestrian Detection [J]. Journal of IEEE Transactions on Systems, Man, andCybernetics,2011,41(1):107-117
[58] Grubb G, Zelinsky A.3D Vision Sensing for Improved Pedestrian Safety [C]//Proceedings of IEEE International Conference on Intelligent Vehicle Symposium, Parma,Italy,2004,19-24
[59]魏闯先.面向行人检测的组合分类器设计[D].合肥：中国科学技术大学，2008
[60] Wu B，Nevatia R．Detection and tracking of multiple，partially occluded humans byBayesian combination of edgelet based part detectors [J]．International Journal ofComputer Vision，2007，75(2)：247-266
[61] Li J，Chua C S．Transductive local exploration particle filter for object tracking[J]．Image and Vision Computing，2007，25(5)：544-552
[62] John L A，Krishnamurthy V．Performance analysis of a dynamic programming trackingbefore detection algorithm [J]． IEEE Transactions on Aerospace and ElectronicSystems，2003，38(1)：228-242
[63]胡洪涛，敬忠良，胡士强．基于辅助粒子滤波的红外小目标检测前跟踪算法[J]．控制与决策，2005，20(11)：1208-1211
[64] Lee M W，Nevatia R．Human pose tracking in monocular sequence using multilevelstructured models [J]．IEEE Transactions on Pattern Analysis and Machine Intelligence，2009，31(1)：27-38
[65] Rogez G，Orrite-Urunuela C，Martinez-Del-Rincon J．A spatio-temporal2D-modelsframework for human pose recovery in monocular sequences [J]．Pattern Recognition，2008，41(9)：2926-2944
[66] Black J，Ellis T．Multi camera image tracking [J]．Image and Vision Computing，2006，24(11)：1256-1267
[67] Khan S M，Shah M．Tracking multiple occluding people by localizing on multiple sceneplanes [J]．IEEE Transactions on Pattern Analysis and Machine Intelligence，2009，31(3)：505-519
[68] Brasnett P，Mihaylova L，Bull D．Sequential Monte Carlo tracking by fusing multiplecues in video sequences [J]．Image and Vision Computing，2007，25(8)：1217-1227
[69] Taycher L，Fisher III J W，Darrell T．Combining object and feature dynamics inprobabilistic tracking [J]．Computer Vision and Image Understanding，2007，108(3)：243-260
[70] CavrilA D M．Multi-cue pedestrian detection and tracking from a moving vehicle[J]．International Journal of Computer Vision，2007，73(1)：41-59
[71] Ommer B，Mader T，Buhmann J M．Seeing the objects behind the dots：recognition invideos from a moving camera [J]．International Journal of Computer Vision，2009，83(1)：57-71
[72]肖念堃，张修如．基于双特征粒子滤波的驾驶员脸部跟踪[J]．计算机系统应用，2008，12：96-98
[73]王建宇，陈熙霖，高文，等．背景变化鲁棒的自适应视觉跟踪模型[J]．软件学报，2006，17(5)：1001-1008
[74] Cheng F H，Chen Y L．Real time multiple objects tracking and identification based ondiscrete wavelet transform [J]．Pattern Recognition，2006，39(6)：1126-1139
[75] Joo S W，Chellappa R．A multiple-hypothesis approach for multiobject visual tracking[J]．IEEE Transactions on Image Processing，2007，16(11)：2849-2854
[76] Briassouli A，Mezaris V，Kompatsiaris I．Color aided motion-segmentation and objecttracking for video sequences semantic analysis [J]．International Journal of ImagingSystems and Technology，2007，17(3)：174-189
[77] Li P H，Chaumette F，Tahri O．A shape tracking algorithm for visual servoing [C]//Proceedings of IEEE International Conference on Robotics and Automation．Barcelona，Spain，2005：2858-2863
[78] Wei F T，Chou S T，Lin C W．A region-based object tracking scheme usingAdaboost-based feature selection [C]//Proceedings of IEEE International Symposiumon Circuits and Systems．Seattle，WA，USA，2008：2753-2758
[79] Pece A E C．Contour tracking based on marginalized likelihood ratios [J]．Image andVision Computing，2006，24(3)：301-317
[80]张旭光，赵恩良，王延杰．基于Mean-shift的灰度目标跟踪新算法[J]．光学技术，2007，33(2)：226-229
[81] Khan Z，Balch T，Dellaert F．MCMC-based particle filtering for tracking a variablenumber of interacting targets [J]．IEEE Transactions on Pattern Analysis and MachineIntelligence，2005，27(11)：1805-1819
[82] Schiff J，Goldberg K．Automated intruder tracking using particle filtering and a networkof binary motion sensors [C]//Proceedings of the IEEE International Conference onAutomation Science and Engineering．Shanghai，China，2006：580-587
[83] Ling J Q，Liu E，Liang H Y．Kernel-based metric for performance evaluation of videoinfrared target tracking [J]．Optical Engineering，2006，45(6)：060505-1-3
[84] Segen Jakub, Kumar, Senthil．Shadow gestures:3D hand pose estimation using a singlecamera [C]//Proceedings of the IEEE Computer Society Conference on ComputerVision and Pattern Recognition，1999，Fort Collins, CO, USA,479-485
[85] Karaulova I.A., Hall P.M., Marshall A.D.．Tracking people in three dimensions using ahierarchical model of dynamics [J]．Image and Vision Computing，2002，20(9)：691-700
[86] Sherrah J., Ristic B., Redding N.J.．Particle filter to track multiple people for visualsurveillance [J]．IET Computer Vision，2011，5(4)：192-200
[87] Stepanov, O.A.． Kalman filtering: Past and present. An outlook from Russia[J]．Gyroscopy and Navigation，2011，2(2)：99-110
[88]李明锁,井亮,邹杰,冯星,姜长生．结合扩展卡尔曼滤波的CamShift移动目标跟踪算法[J]．电光与控制，2011，18(4)：1-5
[89]李鹏,宋申民,陈兴林．自适应平方根无迹卡尔曼滤波算法[J]．控制理论与应用，2010，27(2)：143-146
[90] Sherrah J., Ristic B., Redding N.J.．Particle filter to track multiple people for visualsurveillance [J]．Gyroscopy and Navigation，2011，5(4)：99-200
[91] Xu Kun, Guo Lanying．An anti-occlusion object tracking algorithm based on mean shiftand particle filter [J]．ICIC Express Letters, Part B: Applications，2011，2(1)：95-100
[92] Siradjuddin Indah Agustien, Widyanto M. Rahmat, Basaruddin T.．Particle filter withGaussian weighting for human tracking [J]．Telkomnika，2012，10(6)：1453-1457
[93]张洪建.基于有限集统计学的多目标跟踪算法研究[D].上海：上海交通大学，2009
[94] Abramson Y., Rabinovich A.．Decidability of split equivalence [J]．Information andComputation，2002，173(1)：15-39
[95] Pang Yanwei, Yuan Yuan, Li Xuelong, Pan Jing．Efficient HOG human detection[J]．Signal Processing，2011，91(4)：773-781
[96] Albiol Alberto, Monzo David．Face recognition using HOG-EBGM [J]．PatternRecognition Letters，2008，29(10)：1537-1543
[97] Wu Bo, Nevatia Ram．Detection and tracking of multiple, partially occluded humans byBayesian combination of edgelet based part detectors [J]．International Journal ofComputer Vision，2007，75(2)：247-266
[98] Hu Bin, Wang Shengjin, Ding Xiaoqing．Multi features combination for pedestriandetection [J]．Journal of Multimedia，2010，5(1)：79-84
[99] Zhao Ji, Ma Jiayi, Tian Jinwen, Ma Jie, Zheng Sheng．Boundary extraction usingsupervised edgelet classification [J]．Optical Engineering，2012，51(1)：117-189
[100] Yang Sheng, Liao Xianmei．A pedestrian detection method based on the HOG-LBPfeature and Gentle AdaBoost [J]．International Journal of Advancements in ComputingTechnology，2012，4(19)：553-560
[101]Tamimi Hashem, Andreasson Henrik． Localization of mobile robots withomnidirectional vision using Particle Filter and iterative SIFT [J]．Robotics andAutonomous Systems，2006，54(9)：758-765
[102] Song R., Szymanski J.．Well-distributed SIFT features [J]．Electronics Letters，2009，45(6)：308-310
[103] Windeatt T.．Diversity measures for multiple classifier system analysis and design[J]．Information Fusion，2005，6(1)：21-36
[104] Viola Paul, Jones Michael. Rapid object detection using a boosted cascade of simplefeatures [C]//Proceedings of the IEEE Computer Society Conference on ComputerVision and Pattern Recognition, Kauai, HI, United states，2001：I511-I518
[105] Viola Paul, Jones Michael, Snow, Daniel. Detecting pedestrians using patterns ofmotion and appearance [C]//Proceedings of the IEEE International Conference onComputer Vision, Nice, France,2003：734-741
[106] Chang Hsun-Li, Yue Tai-Wen. Entropy-directed AdaBoost algorithm with NBBPfeatures for face detection [J]． Information Technology Journal,2011,10(8)：1518-1526
[107] Wen Jia, Li Chao, Xiong Zhang. Spatial dependence matrix feature and redundancyelimination algorithm using AdaBoost for object detection [J]． Optical Engineering,2011,50(5)：158-166
[108]邓乃扬,田英杰.支持向量机:理论、算法与拓展[M].北京：科学出版社,2009
[109]肖-泰勒.支持向量机导论[M].北京：电子工业出版社,2004
[110] Drucker Harris, Wu Donghui, Vapnik Vladimir N.. Support vector machines for spamcategorization [J]．IEEE Transactions on Neural Networks,1999,10(5)：1048-1054
[111] Kwok, James Tin-Yau. Moderating the outputs of support vector machine classifiers[J]．IEEE Transactions on Neural Networks,1999,10(5)：1018-1031
[112] Wang Jiaqi, Tao Qing, Wang Jue. Kernel projection algorithm for large-scale SVMproblems [J]．Journal of Computer Science and Technology,2005,17(5)：556-564
[113] Tao Wanga, Zhenxing Qina, Shichao Zhangb, Chengqi Zhanga. Cost-sensitiveclassification with inadequate labeled data [J]．Information Systems,2012,37(5)：508–516
[114] Ana M. Palaciosa, Luciano Sáncheza, Inés Couso. Linguistic cost-sensitive learning ofgenetic fuzzy classifiers for imprecise data [J]．International Journal of ApproximateReasoning,2011,52(6)：841–862
[115] Kuncheva L I, Whitaker C J. Linguistic cost-sensitive learning of genetic fuzzyclassifiers for imprecise data [J]. Measures of diversity in classifier ensembles and theirrelationship with the ensemble accuracy,2003,21(2)：181–207
[116] Collobert R, Sinz F, Weston J. Large scale transductive SVMs [J]. Journal of MachineLearning Research,2006,7(8)：1687–1712
[117] WU B.．Detection of Multiple Partially Occluded Humans in a Single Image byBayesian Combination of Edgelet Part Detectors [C]//IEEE International Conference onComputer Vision,2005,90-97
[118] Valentini G. An experimental bias-variance analysis of SVM ensembles based onresampling techniques [J]. Journal of IEEE Transactions on Systems, Man, andCybernetics,2005,35(6)：1252-1271
[119] Lixia Liu, Yiqi Zhuang, Liu Xueyong. Tax forecasting theory and model based on SVMoptimized by PSO [J]. Expert Systems with Applications,2011,38(1)：116-120
[120]李丽,牛奔.粒子群优化算法[M].北京：冶金工业出版社,2009
[121]纪震,廖惠连,吴青华.粒子群算法及应用[M].北京：科学出版社,2009
[122] Marseguerra, M. A MC-PSO approach to the failure probability evaluation of riskyplant components: The maintenance design [J]. Reliability Engineering and SystemSafety,2013,11(1)：1-8
[123] Jiang Huimin, Kwong C.K., Chen Zengqiang, Ysim Y.C.. Chaos particle swarmoptimization and T-S fuzzy modeling approaches to constrained predictive control [J].Expert Systems with Applications,2012,39(1)：194-201
[124] Tsukamoto Kazuya，Kajikawa Yoshinobu，Nomura Yasuo. Sound field reproductionsystem using simultaneous perturbation method [J]. IEICE Transactions onFundamentals of Electronics, Communications and Computer Sciences,2008, E91-A(3)：801-808
[125] Qi Wu, Rob Law. Cauchy mutation based on objective variable of Gaussian particleswarm optimization for parameters selection of SVM [J]. Expert Systems withApplications,2011,38(2011)：6405–6411
[126] Hanhong Zhu, Yi Wang, Kesheng Wang, Yun Chen. Particle Swarm Optimization (PSO)for the constrained portfolio optimization problem [J]. Expert Systems withApplications,2011,38(2011)：10161–10169
[127] Tang YZ, Li YF．Contour coding based rotating adaptive model for human detectionand tracking in thermal catadioptric omnidirectional vision [J]．APPLIED OPTICS，2012，27(51)：6641-6652
[128] Wang Xin, Tang Zhenmin． Modified particle filter-based infrared pedestriantracking[J]．Infrared Physics and Technology，2010，53(4)：280-287
[129] Kelly Philip， O'Connor Noel E， Smeaton Alan F．Robust pedestrian detection andtracking in crowded scenes [J]．Image and Vision Computing，2009，27(10)：1445-1458
[130] Wang H，Suter D，Sehindler K. Adaptive object tracking based on an effectiveappearance filte [J]．IEEE Trans on Pattern Analysis and Machine Intelligence，2007，9(29)：1661-1667
[131]胡昭华．基于粒子滤波的视频目标跟踪技术研究[D]．南京：南京理工大学，2008
[132] Arulampalam M，Maskell S，Gordon N．A tutorial on particle filters for onlinenonlinear/non-Gaussian Bayesian tracking [J]．IEEE Transactions on Signal Processing，2002，50(2)：174-188
[133]罗飞腾．目标跟踪的粒子滤波技术研究[D]．合肥：中国科学技术大学，2010
[134] Crisan D，Doucet A．A Survey of convergence results on particle filtering methods forpractitioners [J]．IEEE Transactions on Speech and Audio Processing，2002，10(3)：173-185
[135] Doucet A，Freitas J，Gordon N．Sequential Monte Carlo methods in practice [M]．NewYork：Spinger，2001
[136] Doucet A，Godsill S，Andrieu C．On sequential Monte Carlo sampling methods forBayesian filtering [J]．Statistical Computer，2000，10(3)：197-208
[137] Moral D．Measure valued processes and interacting particle systems：Application tononlinear filtering problems [J]．The Annals of Applied Probability，1998，8(2)：438-495
[138] Capellades Martí Balcells, Doermann David, DeMenthon Daniel．An appearance basedapproach for human and object tracking [C]//Proceedings of the IEEE InternationalConference on Image Processing，2003：85-88
[139] Chen Qiang, Sun Quan-Sen, Heng Pheng Ann, Xia De-Shen．Two-stage object trackingmethod based on kernel and active contour [J]．IEEE Transactions on Circuits andSystems for Video Technology，2010，20(4)：605-609
[140] Comaniciu Dorin, Ramesh Visvanathan, Meer Peter．Kernel-based object tracking[J]．IEEE Transactions on Pattern Analysis and Machine Intelligence，2003，25(5)：564-577
[141] Yao Anbang, Lin Xinggang, Wang Guijin, Yu Shan．A compact association of particlefiltering and kernel based object tracking [J]．Pattern Recognition，2012，45(7)：2584-2597
[142] T. Ojala, K. Valkealahti, E. Oja, M. Pietik inen． Texture discrimination withmulti-dimensional distributions of signed gray-level differences [J]． PatternRecognition，2001，34(3)：727–739
[143] T. Ojala, M. Pietik inen, T. M enp. Multiresolution gray-scale and rotation invarianttexture classification with local binary patterns [J]. IEEE Transactions on PatternAnalysis and Machine Intelligence，2006，24(7)：971-987
[144] Porikli F．Integral histogram：a fast way to extract histograms in Cartesian spaces [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition．SanDiego，CA，USA，2005：829-836
[145] Erdem Erkut, Dubuisson Séverine, Bloch Isabelle. Visual tracking by fusing multiplecues with context-sensitive reliabilities [J]. Pattern Recognition，2012，45(5)：1948-1959
[146] Islam Md. Zahidul, Oh Chi-Min, Lee Chil-Woo. An efficient multiple cues synthesis forhuman tracking using a particle filtering framework [J]. International Journal ofInnovative Computing, Information and Control，2011，7(6)：3379-3393
[147] Koora Bharath Kumar, Satpute Nitin R., Adiga, Akshay． Tabu search basedimplementation of object tracking using joint color texture histogram [C]//Proceedingsof IEEE7th International Conference on Industrial and Information Systems．Chennai,India，2012：829-835
[148] Liu Li, Zhao Lingjun, Long Yunli, Kuang Gangyao, Fieguth Paul. Extended local binarypatterns for texture classification [J]. Image and Vision Computing，2012，30(2)：86-99
[149]杨晓明．基于计算机视觉的地铁线网客流实时监控系统研究[D]．北京：北京交通大学，2010
[150] Mohd Asyraf Zulkifley, Bill Moran. Robust hierarchical multiple hypothesis tracker formultiple-object tracking [J]．Expert Systems with Applications，2012，39(2012)：12319–12331
[151] Boyd Jeffrey E., Meloche, Jean. Evaluation of statistical and multiple-hypothesistracking for video traffic surveillance [J]．Machine Vision and Applications，2003，13(5)：344-351
[152] Fuentes L M., Velastin S A. People tracking in surveillance application [J]．Image andVision Computing，2006，24(11)：1165-1171
[153] Panta Kusha, Clark Daniel E., Vo Ba-Ngu. Data association and track management forthe gaussian mixture probability hypothesis density filter [J]．IEEE Transactions onAerospace and Electronic Systems，2009，45(3)：1003-1016
[154]李正周，金钢，董能力.基于改进概率数据关联滤波的红外小目标运动目标跟踪[J]．电子与信息学报，2008，30(4)：953-956
[155] Ding Z., Leung H., Hong, L.. Decoupling joint probabilistic data association algorithmfor multiple target tracking [J]．IEE Proceedings: Radar, Sonar and Navigation，1999，146(5)：251-254
[156] Cham Tat-Jen, Rehg James M.. Multiple hypothesis approach to figure tracking[J]．Proceedings of the IEEE Computer Society Conference on Computer Vision andPattern Recognition，1999，2：239-245
[157] Bar-Shalom Yaakov, Blackman Sam S., Fitzgerald Robert J.. Dimensionless scorefunction for multiple hypothesis tracking [J]．IEEE Transactions on Aerospace andElectronic Systems，2007，43(1)：392-400
[158]韩崇昭，朱洪艳，段战胜.多源信息融合[M].北京：清华大学出版社,2006
[159] Mohd Asyraf Zulkifley, Bill Moran. Robust hierarchical multiple hypothesis tracker formultiple-object tracking [J]．Expert Systems with Applications，2012，39(2012)：12319–12331
[160] Zhang J, Lin J J, Chen X W. Influence of tracking gate upon multiple target tracking[J]．Journal of East China University of Science and Technology，2006，32(12)：1468–1471
[161] R. Danchick, G.E. Newnam. Reformulating Reid's MHT method with generalisedMurty K-best ranked linear assignment algorithm [J]．IEE Proceedings: Radar Sonar andNavigation，2006，153(1)：13–22
[162]刘国成，王永骥.一种基于改进粒子滤波的多目标跟踪算法[J]．控制与决策，2009，24(2)：317-320
[163] Palkki, Ryan D., Lanterman Aaron D., Blair, W. Dale. Addressing Track HypothesisCoalescence in Sequential K-Best Multiple Hypothesis Tracking [J]．IEEE Transactionson Aerospace and Electronic Systems，2011，47(3)：1551-1563
[164]柏柯嘉,刘伟铭,汤义.基于Gabor小波和颜色的阴影检测算法[J]．华南理工大学学报（自然科学版），2009，37(1)：64-68
[2]邹铁方．车人碰撞事故再现及结果不确定性分析[D]．广州：中山大学，2010
[3]申杰．汽车-行人碰撞事故再现研究与应用[D]．上海：上海交通大学，2007
[4] M. Enzweiler, D. M. Gavrila. Monocular pedestrian detection: Survey and experiments[J]．Journal of IEEE Transactions on Pattern Analysis and Machine Intelligence，2009，31(12):2179–2195
[5] Lie Guo, Ping-Shu Ge, Ming-Heng Zhang, et al. Pedestrian detection for intelligenttransportation systems combining AdaBoost algorithm and support vector machine[J]．Expert Systems with Applications，2012，39(4):4274–4286
[6] Ramesh Jain, Rangachar Kasturi, Brian G. Schunck．Machine vision [M]．Bei jing：ChinaMachine Press，2003
[7] P.Viola, M.J.Jones, D.Snow. Detecting Pedestrians Using Patterns of Motion andAppearance [J]. Int.J.Comput.Vision,2005,63(2):153–161
[8]汤义．智能交通系统中基于视频的行人检测与跟踪方法的研究[D]．广州：华南理工大学，2010
[9]田广.基于视觉的行人检测和跟踪技术的研究[D].上海：上海交通大学，2007
[10] Wang Hongliang, Wang Jinqi, Ding Haifei, et al. Moving target detection based on theimproved Gaussian mixture model background difference method [C]//3rd internationalConference on Manufacturing Science and Engineering. Xiamen，China，2012:482-484
[11]朱慧颖，张利，李云廷.基于背景差分法的交通事件智能检测系统[J]．武汉理工大学学报，2011，33(2):79–83
[12]王永忠，梁彦，潘泉，程咏梅，赵春晖.基于自适应混合高斯模型的时空背景建模[J]．自动化学报，2009，35(4):371–378
[13]李娟，邵春福，杨励雅.基于混合高斯模型的行人检测方法[J]．吉林大学学报(工学版)，2011，41(1):41–45
[14] Zhang X, Sexton G. Automatic pedestrian counting using image processing techniques[J]. Electronics Letters,1995,31(11):863–865
[15] Gavrila D.M., Munder S. Multi-cue pedestrian detection and tracking from a movingvehicle [J]. International Journal of Computer Vision,2007,73(1):41-59
[16] Gavrila D.M., Giebel J., Munder S. Vision-based pedestrian detection: ThePROTECTOR system [C]//Proceedings of IEEE Intelligent Vehicles Symposium,Parma, Italy,2004,13-18
[17] Gavrila D.M., Franke U., W hler, C., G rzig, S. Real-time vision for intelligent vehicles[J]. IEEE Instrumentation and Measurement Magazine,2001,4(2):22–27
[18]许言午.面向行人检测的组合分类计算模型与应用研究[D].合肥：中国科学技术大学，2009
[19] Wu Bo, Nevatia Ram. Detection of multiple, partially occluded humans in a single imageby bayesian combination of edgelet part detectors [J] International Journal of ComputerVision,2007,75(2):247-266
[20] Michailovich Oleg, Tannenbaum Allen. A method for prediction and estimation oflarge-amplitude optical flows via extended Kalman filtering approach [J]. EngineeringComputations,2006,23(5):503-514
[21]裴巧娜.基于光流法的运动目标检测与跟踪技术[D].北京：北方工业大学，2009
[22]杨叶梅.基于改进光流法的运动目标检测[J]．计算机与数字工程，2011，39(9):108–111
[23]王春涛.基于背景差分法和光流法的视频动态目标检测与跟踪[J]．软件导报，2011，10(6):145–147
[24] Zimmer Henning, Bruhn Andrés, Weickert Joachim. Optic flow in harmony [J].International Journal of Computer Vision,2011,93(3):368-388
[25]米切尔.机器学习[M].北京：机械工业出版社，2003
[26] Paisitkriangkcrai Sakrapee, Shen Chunhua, Zhang Jian. Fast pedestrian detection using acascade of boosted covariance features [J]. IEEE Transactions on Circuits and Systemsfor Video Technology,2008,18(8):1140-1151
[27] Tuzel Oncel, Porikli Fatih, Meer Peter. Pedestrian detection via classification onRiemannian manifolds [J]. IEEE Transactions on Pattern Analysis And MachineIntelligence,2008,30(10):1713-1727
[28] Viola P., Jones M. Robust real-time face detection [J]. International Journal of ComputerVision,2004,52(2):137-154
[29] Viola P, Jones MJ, Snow D. Detecting pedestrians using patterns of motion andappearance [J]. International Journal of Computer Vision,2005,63(2):153-161
[30] Papageorgiou C, Poggio T. A trainable system for object detection [J]. InternationalJournal of Computer Vision,2000,38(1):15-33
[31] Lienhart Rainer, Maydt Jochen. An extended set of Haar-like features for rapid objectdetection [C]//Proceedings of IEEE International Conference on Image, Rochester, NY,United states,2002,900-903
[32] Xinyi Cui, Yazhou Liu, Shiguang Shan, Xilin Chen, Wen Gao.3D Haar-like features forpedestrian detection [C]//Proceedings of IEEE International Conference on Multimediaand Expo, Beijing, China,2007,1263-1266
[33] Dalal Navneet, Triggs Bill. Histograms of oriented gradients for human detection [C]//Proceedings of IEEE Computer Society Conference on Computer Vision and PatternRecognition, San Diego, CA, United states,2005,886-893
[34] Yangwei Pang, Yuan Yuan, Xuelong Li, Jing Pan. Efficient HOG human detection [J].Signal Processing,2011,91(2011):773-781
[35] Wang Xiaoyu, Han Tony X., Yan Shuicheng. An HOG-LBP Human Detector withPartial Occlusion Handling [C]//Proceedings of12th IEEE International Conference onComputer Vision, Kyoto, JAPAN,2009,32-39
[36] Lowe D M.. Distinctive image features from scale-invariant keypoints [J]. InternationalJournal of Computer Vision,2004,60(2):91-110
[37] Inan Tolga, Halici Ugur.3-D face recognition with local shape descriptors [J]. IEEETransactions on Information Forensics and Security,2012,7(2):577-587
[38]孙玉芬,刘钢.基于SIFT特征的视频内头部姿态估计[J]．武汉理工大学学报，2011，33(1):146–150
[39] Refregier A. Shapelets:I.A method for image analysis [J]. Monthly Notice of the RoyalAstronomical Society,2003,338(1):35-47
[40] Sabzmeydani P, Mori G. Detecting Pedestrians by Learning Shapelet Features [C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition,Minneapolis, MN,2007,1-8
[41] Yang Hee-Deok, Lee Seong-Whan. Multiple pedestrian detection and tracking based onweighted temporal texture features [C]//Proceedings of International Conference onPattern Recognition, British Machine Vis Assoc, Cambridge, ENGLAND,2004,248-251
[42] Zhang Yang, Liu Weiming, Deng Rufeng, Mo Chen. A new pedestrian detection methodbased on diverse AdaBoost-SVM optimization ensemble classifiers algorithm [J].Journal of Computational Information Systems,2011,7(14):5076-5085
[43]苏松志,李绍滋,陈淑媛,蔡国榕,吴云东.行人检测技术综述[J]．电子学报，2012，40(4):814–820
[44]朱文佳.基于机器学习的行人检测关键技术研究[D].上海：上海交通大学，2008
[45] Guo Lie, Ge, Ping-Shu, Zhang Ming-Heng, Li Lin-Hui, Zhao Yi-Bing. Pedestriandetection for intelligent transportation systems combining AdaBoost algorithm andsupport vector machine [J]. Expert Systems with Applications,2012,39(4):4274-4286
[46]李闯,丁晓青,吴佑寿.一种改进的AdaBoost算法—AD AdaBoost [J]．计算机学报，2007，30(1):103–109
[47]张彦峰，何佩琨.一种改进的AdaBoost算法—M-Asy AdaBoost [J]．北京理工大学学报，2011，31(1):64–68
[48]李文辉，倪洪印.一种改进的Adaboost训练算法[J]．吉林大学学报(理学版)，2011，49(3):498–504
[49] J. B′egard N. Allezard, P. Sayd. Real-Time Human Detection in Urban Scenes: LocalDescriptors and Classifiers Selection with AdaBoost-like Algorithms [C]//Proceedingsof IEEE Computer Society Conference on Computer Vision and Pattern RecognitionWorkshops, Anchorage, AK, United states,2008,1-8
[50]朱文佳,戚飞虎.基于Gentle Adaboost的行人检测[J]．中国图象图形学报，2007，12(10):1905–1908
[51] Vapnik V, Chapelle O. Bounds on error expectation for support vector machines [J].Neural Computation,2000,12(9):2013-2036
[52] Ignacio Parra Alonso, David Fernández Llorca, Miguel ángel Sotelo. Combination ofFeature Extraction Methods for SVM Pedestrian Detection [J]. IEEE Transactions onIntelligent Transportation Systems,2007,8(2):292-307
[53]田广,戚飞虎.移动摄像机环境下基于特征变换和SVM的分级行人检测算法[J]．电子学报，2008，36(5):1024–1028
[54] Lee Gyemin, Scott Clayton. Nested support vector machines [J]. IEEE Transactions onSignal Processing,2010,58(3):1648-1660
[55]刘冲,张均东,曾鸿.基于支持向量机的无穷维AdaBoost算法及其应用[J]．仪器仪表学报，2010，31(4):764-769
[56] Strack, Robert, Kecman Vojislav, Strack Beata, Li Qi. Sphere Support Vector Machinesfor large classification tasks [J]. Neurocomputing,2013,101(3):59-67
[57] Xu Y.W., Cao X.B., Qiao H. An Efficient Tree Classifier Ensemble-based Approach forPedestrian Detection [J]. Journal of IEEE Transactions on Systems, Man, andCybernetics,2011,41(1):107-117
[58] Grubb G, Zelinsky A.3D Vision Sensing for Improved Pedestrian Safety [C]//Proceedings of IEEE International Conference on Intelligent Vehicle Symposium, Parma,Italy,2004,19-24
[59]魏闯先.面向行人检测的组合分类器设计[D].合肥：中国科学技术大学，2008
[60] Wu B，Nevatia R．Detection and tracking of multiple，partially occluded humans byBayesian combination of edgelet based part detectors [J]．International Journal ofComputer Vision，2007，75(2)：247-266
[61] Li J，Chua C S．Transductive local exploration particle filter for object tracking[J]．Image and Vision Computing，2007，25(5)：544-552
[62] John L A，Krishnamurthy V．Performance analysis of a dynamic programming trackingbefore detection algorithm [J]． IEEE Transactions on Aerospace and ElectronicSystems，2003，38(1)：228-242
[63]胡洪涛，敬忠良，胡士强．基于辅助粒子滤波的红外小目标检测前跟踪算法[J]．控制与决策，2005，20(11)：1208-1211
[64] Lee M W，Nevatia R．Human pose tracking in monocular sequence using multilevelstructured models [J]．IEEE Transactions on Pattern Analysis and Machine Intelligence，2009，31(1)：27-38
[65] Rogez G，Orrite-Urunuela C，Martinez-Del-Rincon J．A spatio-temporal2D-modelsframework for human pose recovery in monocular sequences [J]．Pattern Recognition，2008，41(9)：2926-2944
[66] Black J，Ellis T．Multi camera image tracking [J]．Image and Vision Computing，2006，24(11)：1256-1267
[67] Khan S M，Shah M．Tracking multiple occluding people by localizing on multiple sceneplanes [J]．IEEE Transactions on Pattern Analysis and Machine Intelligence，2009，31(3)：505-519
[68] Brasnett P，Mihaylova L，Bull D．Sequential Monte Carlo tracking by fusing multiplecues in video sequences [J]．Image and Vision Computing，2007，25(8)：1217-1227
[69] Taycher L，Fisher III J W，Darrell T．Combining object and feature dynamics inprobabilistic tracking [J]．Computer Vision and Image Understanding，2007，108(3)：243-260
[70] CavrilA D M．Multi-cue pedestrian detection and tracking from a moving vehicle[J]．International Journal of Computer Vision，2007，73(1)：41-59
[71] Ommer B，Mader T，Buhmann J M．Seeing the objects behind the dots：recognition invideos from a moving camera [J]．International Journal of Computer Vision，2009，83(1)：57-71
[72]肖念堃，张修如．基于双特征粒子滤波的驾驶员脸部跟踪[J]．计算机系统应用，2008，12：96-98
[73]王建宇，陈熙霖，高文，等．背景变化鲁棒的自适应视觉跟踪模型[J]．软件学报，2006，17(5)：1001-1008
[74] Cheng F H，Chen Y L．Real time multiple objects tracking and identification based ondiscrete wavelet transform [J]．Pattern Recognition，2006，39(6)：1126-1139
[75] Joo S W，Chellappa R．A multiple-hypothesis approach for multiobject visual tracking[J]．IEEE Transactions on Image Processing，2007，16(11)：2849-2854
[76] Briassouli A，Mezaris V，Kompatsiaris I．Color aided motion-segmentation and objecttracking for video sequences semantic analysis [J]．International Journal of ImagingSystems and Technology，2007，17(3)：174-189
[77] Li P H，Chaumette F，Tahri O．A shape tracking algorithm for visual servoing [C]//Proceedings of IEEE International Conference on Robotics and Automation．Barcelona，Spain，2005：2858-2863
[78] Wei F T，Chou S T，Lin C W．A region-based object tracking scheme usingAdaboost-based feature selection [C]//Proceedings of IEEE International Symposiumon Circuits and Systems．Seattle，WA，USA，2008：2753-2758
[79] Pece A E C．Contour tracking based on marginalized likelihood ratios [J]．Image andVision Computing，2006，24(3)：301-317
[80]张旭光，赵恩良，王延杰．基于Mean-shift的灰度目标跟踪新算法[J]．光学技术，2007，33(2)：226-229
[81] Khan Z，Balch T，Dellaert F．MCMC-based particle filtering for tracking a variablenumber of interacting targets [J]．IEEE Transactions on Pattern Analysis and MachineIntelligence，2005，27(11)：1805-1819
[82] Schiff J，Goldberg K．Automated intruder tracking using particle filtering and a networkof binary motion sensors [C]//Proceedings of the IEEE International Conference onAutomation Science and Engineering．Shanghai，China，2006：580-587
[83] Ling J Q，Liu E，Liang H Y．Kernel-based metric for performance evaluation of videoinfrared target tracking [J]．Optical Engineering，2006，45(6)：060505-1-3
[84] Segen Jakub, Kumar, Senthil．Shadow gestures:3D hand pose estimation using a singlecamera [C]//Proceedings of the IEEE Computer Society Conference on ComputerVision and Pattern Recognition，1999，Fort Collins, CO, USA,479-485
[85] Karaulova I.A., Hall P.M., Marshall A.D.．Tracking people in three dimensions using ahierarchical model of dynamics [J]．Image and Vision Computing，2002，20(9)：691-700
[86] Sherrah J., Ristic B., Redding N.J.．Particle filter to track multiple people for visualsurveillance [J]．IET Computer Vision，2011，5(4)：192-200
[87] Stepanov, O.A.． Kalman filtering: Past and present. An outlook from Russia[J]．Gyroscopy and Navigation，2011，2(2)：99-110
[88]李明锁,井亮,邹杰,冯星,姜长生．结合扩展卡尔曼滤波的CamShift移动目标跟踪算法[J]．电光与控制，2011，18(4)：1-5
[89]李鹏,宋申民,陈兴林．自适应平方根无迹卡尔曼滤波算法[J]．控制理论与应用，2010，27(2)：143-146
[90] Sherrah J., Ristic B., Redding N.J.．Particle filter to track multiple people for visualsurveillance [J]．Gyroscopy and Navigation，2011，5(4)：99-200
[91] Xu Kun, Guo Lanying．An anti-occlusion object tracking algorithm based on mean shiftand particle filter [J]．ICIC Express Letters, Part B: Applications，2011，2(1)：95-100
[92] Siradjuddin Indah Agustien, Widyanto M. Rahmat, Basaruddin T.．Particle filter withGaussian weighting for human tracking [J]．Telkomnika，2012，10(6)：1453-1457
[93]张洪建.基于有限集统计学的多目标跟踪算法研究[D].上海：上海交通大学，2009
[94] Abramson Y., Rabinovich A.．Decidability of split equivalence [J]．Information andComputation，2002，173(1)：15-39
[95] Pang Yanwei, Yuan Yuan, Li Xuelong, Pan Jing．Efficient HOG human detection[J]．Signal Processing，2011，91(4)：773-781
[96] Albiol Alberto, Monzo David．Face recognition using HOG-EBGM [J]．PatternRecognition Letters，2008，29(10)：1537-1543
[97] Wu Bo, Nevatia Ram．Detection and tracking of multiple, partially occluded humans byBayesian combination of edgelet based part detectors [J]．International Journal ofComputer Vision，2007，75(2)：247-266
[98] Hu Bin, Wang Shengjin, Ding Xiaoqing．Multi features combination for pedestriandetection [J]．Journal of Multimedia，2010，5(1)：79-84
[99] Zhao Ji, Ma Jiayi, Tian Jinwen, Ma Jie, Zheng Sheng．Boundary extraction usingsupervised edgelet classification [J]．Optical Engineering，2012，51(1)：117-189
[100] Yang Sheng, Liao Xianmei．A pedestrian detection method based on the HOG-LBPfeature and Gentle AdaBoost [J]．International Journal of Advancements in ComputingTechnology，2012，4(19)：553-560
[101]Tamimi Hashem, Andreasson Henrik． Localization of mobile robots withomnidirectional vision using Particle Filter and iterative SIFT [J]．Robotics andAutonomous Systems，2006，54(9)：758-765
[102] Song R., Szymanski J.．Well-distributed SIFT features [J]．Electronics Letters，2009，45(6)：308-310
[103] Windeatt T.．Diversity measures for multiple classifier system analysis and design[J]．Information Fusion，2005，6(1)：21-36
[104] Viola Paul, Jones Michael. Rapid object detection using a boosted cascade of simplefeatures [C]//Proceedings of the IEEE Computer Society Conference on ComputerVision and Pattern Recognition, Kauai, HI, United states，2001：I511-I518
[105] Viola Paul, Jones Michael, Snow, Daniel. Detecting pedestrians using patterns ofmotion and appearance [C]//Proceedings of the IEEE International Conference onComputer Vision, Nice, France,2003：734-741
[106] Chang Hsun-Li, Yue Tai-Wen. Entropy-directed AdaBoost algorithm with NBBPfeatures for face detection [J]． Information Technology Journal,2011,10(8)：1518-1526
[107] Wen Jia, Li Chao, Xiong Zhang. Spatial dependence matrix feature and redundancyelimination algorithm using AdaBoost for object detection [J]． Optical Engineering,2011,50(5)：158-166
[108]邓乃扬,田英杰.支持向量机:理论、算法与拓展[M].北京：科学出版社,2009
[109]肖-泰勒.支持向量机导论[M].北京：电子工业出版社,2004
[110] Drucker Harris, Wu Donghui, Vapnik Vladimir N.. Support vector machines for spamcategorization [J]．IEEE Transactions on Neural Networks,1999,10(5)：1048-1054
[111] Kwok, James Tin-Yau. Moderating the outputs of support vector machine classifiers[J]．IEEE Transactions on Neural Networks,1999,10(5)：1018-1031
[112] Wang Jiaqi, Tao Qing, Wang Jue. Kernel projection algorithm for large-scale SVMproblems [J]．Journal of Computer Science and Technology,2005,17(5)：556-564
[113] Tao Wanga, Zhenxing Qina, Shichao Zhangb, Chengqi Zhanga. Cost-sensitiveclassification with inadequate labeled data [J]．Information Systems,2012,37(5)：508–516
[114] Ana M. Palaciosa, Luciano Sáncheza, Inés Couso. Linguistic cost-sensitive learning ofgenetic fuzzy classifiers for imprecise data [J]．International Journal of ApproximateReasoning,2011,52(6)：841–862
[115] Kuncheva L I, Whitaker C J. Linguistic cost-sensitive learning of genetic fuzzyclassifiers for imprecise data [J]. Measures of diversity in classifier ensembles and theirrelationship with the ensemble accuracy,2003,21(2)：181–207
[116] Collobert R, Sinz F, Weston J. Large scale transductive SVMs [J]. Journal of MachineLearning Research,2006,7(8)：1687–1712
[117] WU B.．Detection of Multiple Partially Occluded Humans in a Single Image byBayesian Combination of Edgelet Part Detectors [C]//IEEE International Conference onComputer Vision,2005,90-97
[118] Valentini G. An experimental bias-variance analysis of SVM ensembles based onresampling techniques [J]. Journal of IEEE Transactions on Systems, Man, andCybernetics,2005,35(6)：1252-1271
[119] Lixia Liu, Yiqi Zhuang, Liu Xueyong. Tax forecasting theory and model based on SVMoptimized by PSO [J]. Expert Systems with Applications,2011,38(1)：116-120
[120]李丽,牛奔.粒子群优化算法[M].北京：冶金工业出版社,2009
[121]纪震,廖惠连,吴青华.粒子群算法及应用[M].北京：科学出版社,2009
[122] Marseguerra, M. A MC-PSO approach to the failure probability evaluation of riskyplant components: The maintenance design [J]. Reliability Engineering and SystemSafety,2013,11(1)：1-8
[123] Jiang Huimin, Kwong C.K., Chen Zengqiang, Ysim Y.C.. Chaos particle swarmoptimization and T-S fuzzy modeling approaches to constrained predictive control [J].Expert Systems with Applications,2012,39(1)：194-201
[124] Tsukamoto Kazuya，Kajikawa Yoshinobu，Nomura Yasuo. Sound field reproductionsystem using simultaneous perturbation method [J]. IEICE Transactions onFundamentals of Electronics, Communications and Computer Sciences,2008, E91-A(3)：801-808
[125] Qi Wu, Rob Law. Cauchy mutation based on objective variable of Gaussian particleswarm optimization for parameters selection of SVM [J]. Expert Systems withApplications,2011,38(2011)：6405–6411
[126] Hanhong Zhu, Yi Wang, Kesheng Wang, Yun Chen. Particle Swarm Optimization (PSO)for the constrained portfolio optimization problem [J]. Expert Systems withApplications,2011,38(2011)：10161–10169
[127] Tang YZ, Li YF．Contour coding based rotating adaptive model for human detectionand tracking in thermal catadioptric omnidirectional vision [J]．APPLIED OPTICS，2012，27(51)：6641-6652
[128] Wang Xin, Tang Zhenmin． Modified particle filter-based infrared pedestriantracking[J]．Infrared Physics and Technology，2010，53(4)：280-287
[129] Kelly Philip， O'Connor Noel E， Smeaton Alan F．Robust pedestrian detection andtracking in crowded scenes [J]．Image and Vision Computing，2009，27(10)：1445-1458
[130] Wang H，Suter D，Sehindler K. Adaptive object tracking based on an effectiveappearance filte [J]．IEEE Trans on Pattern Analysis and Machine Intelligence，2007，9(29)：1661-1667
[131]胡昭华．基于粒子滤波的视频目标跟踪技术研究[D]．南京：南京理工大学，2008
[132] Arulampalam M，Maskell S，Gordon N．A tutorial on particle filters for onlinenonlinear/non-Gaussian Bayesian tracking [J]．IEEE Transactions on Signal Processing，2002，50(2)：174-188
[133]罗飞腾．目标跟踪的粒子滤波技术研究[D]．合肥：中国科学技术大学，2010
[134] Crisan D，Doucet A．A Survey of convergence results on particle filtering methods forpractitioners [J]．IEEE Transactions on Speech and Audio Processing，2002，10(3)：173-185
[135] Doucet A，Freitas J，Gordon N．Sequential Monte Carlo methods in practice [M]．NewYork：Spinger，2001
[136] Doucet A，Godsill S，Andrieu C．On sequential Monte Carlo sampling methods forBayesian filtering [J]．Statistical Computer，2000，10(3)：197-208
[137] Moral D．Measure valued processes and interacting particle systems：Application tononlinear filtering problems [J]．The Annals of Applied Probability，1998，8(2)：438-495
[138] Capellades Martí Balcells, Doermann David, DeMenthon Daniel．An appearance basedapproach for human and object tracking [C]//Proceedings of the IEEE InternationalConference on Image Processing，2003：85-88
[139] Chen Qiang, Sun Quan-Sen, Heng Pheng Ann, Xia De-Shen．Two-stage object trackingmethod based on kernel and active contour [J]．IEEE Transactions on Circuits andSystems for Video Technology，2010，20(4)：605-609
[140] Comaniciu Dorin, Ramesh Visvanathan, Meer Peter．Kernel-based object tracking[J]．IEEE Transactions on Pattern Analysis and Machine Intelligence，2003，25(5)：564-577
[141] Yao Anbang, Lin Xinggang, Wang Guijin, Yu Shan．A compact association of particlefiltering and kernel based object tracking [J]．Pattern Recognition，2012，45(7)：2584-2597
[142] T. Ojala, K. Valkealahti, E. Oja, M. Pietik inen． Texture discrimination withmulti-dimensional distributions of signed gray-level differences [J]． PatternRecognition，2001，34(3)：727–739
[143] T. Ojala, M. Pietik inen, T. M enp. Multiresolution gray-scale and rotation invarianttexture classification with local binary patterns [J]. IEEE Transactions on PatternAnalysis and Machine Intelligence，2006，24(7)：971-987
[144] Porikli F．Integral histogram：a fast way to extract histograms in Cartesian spaces [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition．SanDiego，CA，USA，2005：829-836
[145] Erdem Erkut, Dubuisson Séverine, Bloch Isabelle. Visual tracking by fusing multiplecues with context-sensitive reliabilities [J]. Pattern Recognition，2012，45(5)：1948-1959
[146] Islam Md. Zahidul, Oh Chi-Min, Lee Chil-Woo. An efficient multiple cues synthesis forhuman tracking using a particle filtering framework [J]. International Journal ofInnovative Computing, Information and Control，2011，7(6)：3379-3393
[147] Koora Bharath Kumar, Satpute Nitin R., Adiga, Akshay． Tabu search basedimplementation of object tracking using joint color texture histogram [C]//Proceedingsof IEEE7th International Conference on Industrial and Information Systems．Chennai,India，2012：829-835
[148] Liu Li, Zhao Lingjun, Long Yunli, Kuang Gangyao, Fieguth Paul. Extended local binarypatterns for texture classification [J]. Image and Vision Computing，2012，30(2)：86-99
[149]杨晓明．基于计算机视觉的地铁线网客流实时监控系统研究[D]．北京：北京交通大学，2010
[150] Mohd Asyraf Zulkifley, Bill Moran. Robust hierarchical multiple hypothesis tracker formultiple-object tracking [J]．Expert Systems with Applications，2012，39(2012)：12319–12331
[151] Boyd Jeffrey E., Meloche, Jean. Evaluation of statistical and multiple-hypothesistracking for video traffic surveillance [J]．Machine Vision and Applications，2003，13(5)：344-351
[152] Fuentes L M., Velastin S A. People tracking in surveillance application [J]．Image andVision Computing，2006，24(11)：1165-1171
[153] Panta Kusha, Clark Daniel E., Vo Ba-Ngu. Data association and track management forthe gaussian mixture probability hypothesis density filter [J]．IEEE Transactions onAerospace and Electronic Systems，2009，45(3)：1003-1016
[154]李正周，金钢，董能力.基于改进概率数据关联滤波的红外小目标运动目标跟踪[J]．电子与信息学报，2008，30(4)：953-956
[155] Ding Z., Leung H., Hong, L.. Decoupling joint probabilistic data association algorithmfor multiple target tracking [J]．IEE Proceedings: Radar, Sonar and Navigation，1999，146(5)：251-254
[156] Cham Tat-Jen, Rehg James M.. Multiple hypothesis approach to figure tracking[J]．Proceedings of the IEEE Computer Society Conference on Computer Vision andPattern Recognition，1999，2：239-245
[157] Bar-Shalom Yaakov, Blackman Sam S., Fitzgerald Robert J.. Dimensionless scorefunction for multiple hypothesis tracking [J]．IEEE Transactions on Aerospace andElectronic Systems，2007，43(1)：392-400
[158]韩崇昭，朱洪艳，段战胜.多源信息融合[M].北京：清华大学出版社,2006
[159] Mohd Asyraf Zulkifley, Bill Moran. Robust hierarchical multiple hypothesis tracker formultiple-object tracking [J]．Expert Systems with Applications，2012，39(2012)：12319–12331
[160] Zhang J, Lin J J, Chen X W. Influence of tracking gate upon multiple target tracking[J]．Journal of East China University of Science and Technology，2006，32(12)：1468–1471
[161] R. Danchick, G.E. Newnam. Reformulating Reid's MHT method with generalisedMurty K-best ranked linear assignment algorithm [J]．IEE Proceedings: Radar Sonar andNavigation，2006，153(1)：13–22
[162]刘国成，王永骥.一种基于改进粒子滤波的多目标跟踪算法[J]．控制与决策，2009，24(2)：317-320
[163] Palkki, Ryan D., Lanterman Aaron D., Blair, W. Dale. Addressing Track HypothesisCoalescence in Sequential K-Best Multiple Hypothesis Tracking [J]．IEEE Transactionson Aerospace and Electronic Systems，2011，47(3)：1551-1563
[164]柏柯嘉,刘伟铭,汤义.基于Gabor小波和颜色的阴影检测算法[J]．华南理工大学学报（自然科学版），2009，37(1)：64-68