摘要
目前我国大多数的煤矿视频监控系统还主要停留在人工监控阶段，智能化煤矿视频监控系统是发展的必然趋势。它可以自动采集获得视频监控图像序列，进行实时运动目标检测、识别和跟踪，通过理解分析图像画面主动发现违规行为、可疑目标和潜在危险，以快速合理的方式发出警报，指导启动相应的联动控制措施。煤矿智能视频监控系统的实现，需要综合运用图像处理、机器学习和计算机视觉等领域中的多项技术，本文对其中的四类关键技术进行研究，具体工作包括：
为了对伴有随机噪声的煤矿雾尘图像进行清晰化处理，提出一种基于DCPBF的去雾除尘和同步去噪算法。推导建立煤矿雾尘降质图像退化模型；设计基于暗原色先验知识的环境光、粗略透射率估计方法与步骤；采用联合双边滤波快速获得精细透射率图；依据图像退化模型构建正则化目标函数，求取转换图像并进行高斯双边滤波，获得去雾除尘图像且同步实现噪声的有效去除。
针对相对静止的煤矿视频监控环境背景，采用背景减除法进行运动目标检测。提出基于聚类技术的自适应背景建模与更新方法，利用改进的FCM算法对像素灰度取值进行聚类，自适应选取不同个数的聚类构建各像素背景模型，随场景变化进行聚类修改、添加和删除完成背景更新。联合背景差分信息、三帧差分信息和空间邻域信息进行前景检测，通过改进的OTSU方法自动设置差分阈值。提出结合像素亮度和纹理特征的运动阴影检测方法，依据在阴影覆盖前后的灰度图像中，像素具有亮度值相关性和纹理特征值不变性，实现运动阴影的检测与去除。
将单目标跟踪看作为目标和背景的在线分类问题，选用线性SVM作为分类工具，提出一种添加样本约简机制的FLSVMIL方法实现分类器在线更新，并提出基于FLSVMIL的单目标跟踪算法。由于可能受到无效历史信息的干扰，并且难以处理样本集非线性可分的问题，提出基于LSVMSE的单目标跟踪算法，采用集成分类器进行运动目标跟踪。
根据煤矿智能视频监控系统中多目标跟踪的任务需求，提出基于UKF-MHT的多目标跟踪算法。设计算法的基本框架，确定关键步骤的处理方法，其中包括跟踪门设置、目标预测值与观测值的数据匹配、航迹评价与删除、航迹聚类和m-best假设的产生以及目标状态的预测更新。在自适应跟踪修正阶段，针对由目标短暂丢失、粘连和分裂可能引起的三类跟踪错误，设计具体的判别策略和修正方法。
At present most of our coalmine video surveillance system is still at the stage ofmanual monitoring, intelligent system is an inevitable trend of development. It canautomatically capture surveillance video image sequence, detect, identificate andtrack moving targets in real time. By analyzing image screen, the intelligent systemcan proactively uncover violations, suspicious targets and potential hazards.And thenquickly alerting by a reasonable way, it can guide the activation of correspondinglinkage controls. The implementation of coalmine intelligent video surveillancesystem requires using a number of technologies such as image processing, machinelearning and computer vision and so on. This paper studies four key technologies; itsspecific tasks are as follows:
In order to clear coal fog dust images with random noise, an algorithm of fogdust removal and simultaneously denoising based on DCPBF is proposed. This paperestablishes a coalmine fog dust image degradation model, designs methods andprocedures for estimating ambient light and rough transmittance based on darkchannel prior. The fine transmittance diagram is quickly obtained by joint bilateralfiltering. A regularization objective function is constructed based on the imagedegradation model. By solving a converted image and Gaussian bilateral filtering theimage, fog dust removal and simultaneously denoising are realized.
Aiming at the relatively static background of coalmine video surveillanceenvironment, this paper uses background subtraction method for moving targetdetection. An adaptive background modeling and updating method based onclustering technology is proposed. Clustering pixel gray values by improved FCMalgorithm, a different number of classifications are adaptively selected to build thebackground model of each pixel. With scene changes classifications are updated,added and deleted thus completing background update. Image foreground is detectedby jointing background differential information, three differential information andspatial information. Differential threshold is automatically setted by improved OTSUmethod. A moving shadow detection method using pixel luminance and texture isproposed. Because the pixel brightness and texture value are invariant in gray imagesbefore and after shadow covering, moving shadow detection and removal can berealized.
Considerring single-target tracking as an online classification problem between target an background, a linear SVM is used as a classification tool. An FLSVMILmethod with sample reduction mechanism is proposed to realize online updatingclassifiers. And so an single-target tracking algorithm based on FLSVMILis proposed.Due to the interference of invalid history information and the nonlinear separability ofsample set, an single-target tracking algorithm based on LSVMSE is proposed.Moving target is tracking by an ensemble classifier.
According to the requirement of multi-target tracking mission in the coalmineintelligent video surveillance system, an multi-target tracking algorithm based onUKF-MHT is proposed. This paper designs the basic algorithm framework, anddetermines the treatment of critical steps which include setting tracking gate,matching target predicted value and observed value, evaluating and removing track,clustering track and generating m-best hypotheses, predicting and updating targetstates. In the process of adaptive tracking correction, specific discriminant strategiesand correction method are designed for three types of tracking error caused by targettemporary loss, target adhesions and split.
In this paper, there are fifty-seven figures, twenty-one tables, and one hundredand fifty-eight reference documents.
引文
[1]张洪杰.煤矿安全风险综合评价体系及应用研究[D].徐州：中国矿业大学，2010.
[2]孙继平.煤矿安全生产监控与通信技术[M].北京：煤炭工业出版社，2009：15-16.
[3]厉丹.视频目标检测与跟踪算法及其在煤矿中应用的研究[D].徐州：中国矿业大学，2011.
[4]周磊.基于注意机制的煤矿监控图像知觉编组研究[D].徐州：中国矿业大学，2010.
[5]袁国武.智能视频监控中的运动目标检测和跟踪算法研究[D].昆明:云南大学博士学位论文,2012.
[6]禹晶，徐东彬，廖庆敏.图像去雾技术研究进展[J].中国图象图形学报，2011,16(9)：1561-1574.
[7]Oakley J P,Bu H.Correction of simple contrast loss in clor images[J].IEEE Transactions onImage Processing,2012,16(2):511-522.
[8]许志远.雾天降质图像增强方法研究及DSP实现[D].大连：大连海事大学，2010.
[9]Stark J A.Adaptive image contrast enhancement using generalizations of histogramequalization[J].IEEE Transactions on Image Processing,2012,9(5):889-896.
[10]Kim J Y,Kim L S,Hwang S H.An advanced contrast enhancement using partiallyoverlapped sub-block histogram equalization[J].Circuits and Systems for VideoTechnology,2009,11(4):475-484.
[11]Seow M-J,Asari V K.Ratio rule and homomorphic filter for enhancement of digital colourimage[J].Neurocomputing,2012,69(7):954-958.
[12]Elad M,Kimmel R,Shaked D.Reduced complexity Retinex algorithm via the variationalapproach[J].Journal of Visual Communication and Image Representation,2008,14(4):36-38.
[13]Provenzi E,Fierro M,Rizzi A.Random spray retinex:a new Retinex implementation toinvestigate the local properties of the model[J].IEEE Transactions on ImageProcessing,2011,16(1):162-171.
[14]Jisha John and M.Wilscy.Enhancement of Weather Degraded Video Sequences UsingWavelet Fusion[C].The7th IEEE International Conference on Cybernnetic IntelligentSystems,2012,64-67.
[15]Eriksson B.Automatic image de-weathering using curvelet-based vanishing pointdetection[EB/OL].[2012-01-20].http://homepages.cae.wisc.edu/~beriksso/cs766.pdf.
[16]郭珈，王孝通，胡程鹏，等.基于单幅图像景深和大气散射模型的去雾方法[J].中国图象图形学报，2012，17(1):27-32.
[17]Kopf J,Neubert B,Chen B.Deep photo:Model-base photograph enhancement andviewing[J].Acm Transanctions on Graphics,2012,27(5):1160-1170.
[18]Schechner Y Y,Averbuch Y.Regularized image recovery in scattering media[J].IEEETransactions on PatternAnalysis and Machine Intelligence,2010,29(9):1655-1660.
[19] Narasim han S G,Nayar S K. Interaetive weathering of anim age using physieal models[C].ICCV Workshop on Color and Photometric Methods in Computer vision. NewYork, USA,2012:1-8.
[20]禹晶，李大鹏，廖庆敏.基于物理模型的快速单幅图像去雾方法[J].自动化学报，2011,37（2）：143-149.
[21]王斌，肖文华，张茂军，等.采用时空条件信息的动态场景运动目标检测[J].计算机辅助设计与图形学学报，2012，24(12):1576-1584.
[22] RadguiA, Demonceaux C, Mouaddib E.Optical flow estimation from multichannel sphericalimage decomposition[J]. Computer Vision and Image Understanding,2012,115(9):12631272.
[23]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.
[24]Brutzer S,Hoferlin B,Heidemann G.Evaluation of background subtraction techniques videosureillance [C].2012IEEE Conference on Computer Vision and Pattern Recognition.Providence:IEEE,2012:1937-1944.
[25]SanMiguel J C,Maritinez J M.On the evaluation of background subtraction algorithms withoutgroundtruth[C].2011Seventh IEEE International Conference onAdvanced Video and SignalBased Surveillance.Boston:IEEE,2011:180-187.
[26]孙春凤.基于并行处理的高速图像序列运动目标检测技术研究[D].哈尔滨：哈尔滨工业大学，2010.
[27]Bin Q,Ghazal M,AmerA.Robust global motion estimation oriented to video objectsegmentation[J].IEEE Transactions on Image Processing,2008,17(6):958-967.
[28]王爱平.视频目标跟踪技术研究[D].长沙：国防科学技术大学，2011.
[29]Babenko B,Yang M H, Belongie S. Robust object tracking with online multiple instancelearning[J]. IEEE Transactions on PatternAnalysis and Machine Intelligence,2011,33(8):16191632.
[30]Han Z, Ye Q, Jiao J. Combined feature evaluation for adaptive visual object tracking[J].Computer Vision and Image Understanding,2011,115(1):6980.
[31]Khan Z H, Gu IY, BackhouseAG. Robust visual object tracking using multi-modeanisotropic mean shift and particle filters[J]. IEEE Transactions on Circuits and Systems for VideoTechnology,2012,21(1):7487.
[32]YilmazA，Javed O,Shah M. Object tracking: a survey［J］.ACM Computing Surveys,2008,38(4):13-45.
[33]Murshed M,Kabir M.H,Chae O.Moving object tracking-an edge segment based approach[J].International Journal of Innovative Computing,2012,22(7):3963-3979.
[34]胡燕.基于GPU并行计算的目标跟踪快速算法研究[D].华中师范大学，2012.
[35]JufengGe,YuPinLuo,GyomeTei,Real-time Pedestrian deteetion and tracking at nighttime fordriver-assistance systems[J].IEEE Transaetionson Intelligent Transportation systems,201110(2):283一298.
[36]Wu M J, Peng X R, Zhang Q H, et al. Patches-based Markov random field model for multipleobject tracking under occlusion[J]. Signal Processing,2012,90(5):1518-1529.
[37] Mikolajczyk K, Tuytelaars T, Schmid C,et al.Acomparison of affine region detectors[J].International Journal of Computer Vision,2012,65(1/2):43-72.
[38]Xu L Q, Landabaso J L, Lei B. Segmentation and tracking of multiple moving objects forintelligent video analysis[J]. BT Technology Journal,2011,22(3):140-150.
[39]Polat E,Yeasin M, Sharma R.A2D/3D model-based object tracking framework[J]. PatternRecognition,2008,36(9):2127-2141.
[40] Li X, Hu W M, Wang H Z. Robust object tracking using a spatial pyramid heat kernelstructural information representation[J]. Neurocomputing,2012,73(17):3179-3190.
[41] Paragios N, Deriche R. Geodesic active regions and level set metnods for motion estimationand tracking[J]. Computer Vision and Image Understanding,2011,97(3):259-282.
[42] Mikolajczyk K, Schmid C. APerformance Evaluation of Local Descriptors[J]. IEEETransaction on PatternAnalysis and Machine Intelligence,2009,27(10):1615-1630.
[43]徐建军，张蓉，毕笃彦，等.一种新的Adaboost视频跟踪算法[J].控制与决策,2012,27(5):681-684.
[44]Okuma1K，TaleghaniA.Aboosted particle filter：Multitarget detection and tracking[C].Proceedings of ECCV2009：28-39
[45] COLLINS R T，LIU Yan-xi．On-line selection of discriminative trac-king features[J]．IEEETransaction on PatternAnalysis and Machine Intelligence，2011，27(10):1631-1643．
[46]Avidan S.Ensemble tracking[J].IEEE Transaction on PatternAnalysis and MachineIntelligence,2011,29(2):261-271.
[47]宋野，齐志泉，王来生.多事例在线学习方法在遮挡目标跟踪中的应用[J].中南大学学报，2011，42（1）：666-670.
[48]杨金龙.被动多传感器目标跟踪及航迹维持算法研究[D].西安:西安电子科技大学，2012.
[49]权太范.目标跟踪新理论与技术[M].北京:国防工业出版社,2009.
[50] X. Yi,Y. He, X. Guan. Cooperative Location Model under the Nearest Neighbor Criterion[C].IEEE PLANS, Monterey, CA, USA,2010:658~661
[51]Samuel,Robert.Design andAnalysis of Modern Tracking Systems[M].London:ArtechHouse,2008.
[52] S. B. Colegrove, S. J. Davey. PDAF with Multiple Clutter Regions and Target Models[J].IEEE Transactions onAerospace and Electronic Systems,2011,39(2):110~123
[53] S. Puranik, J. K. Tugnait. Tracking of Multiple Maneuvering Targets Using Multiscan JPDAand IMM Filtering[J]. IEEE Transactions onAerospace and Electronic Systems,2011,43(1):23~35.
[54]James L,Fisher,David P.Fast JPDA multitarget tracking algorithm[J].Applied Optics,2008,28(2):371-378.
[55]Kennedy H.L.Controlling track coalescence with scaled Joint Probabilistic DataAssociation[J].IEEE Radar,2010,32(8):440-445.
[56]Song-lin Chen,Yi-bing Xu.ANew Joint Possibility DataAssociationAlgorithmAvoidingTrack Coalescence[J].Intelligent Systems andApplications,2011,34(2):45-51.
[57] S. S. Blackman. Multiple Hypothesis Tracking for Multiple Target Tracking[J].IEEEAerospace and Electronic Systems.2008,19(1):5~18.
[58]BAR-SHALOMY.Tracking with classification-aided multiframe data association[J].IEEETransaction onAerospace and Electronics Systems,2008,41(3):868-878.
[59]AsadA,Robert T. Multi-Target Tracking by Lagrangian Relaxation to Min-Cost NetworkFlow[C].CVPR2013,Oregon,USA,2013:2614-2621.
[60]李宏博.高频雷达目标数据处理技术研究[D].哈尔滨：哈尔滨工业大学，2009.
[61] Vo B-N,Singh S,DoucetA.Sequential Monte Carloimplementation of the PHD filter formulti-target track-ing[C].Proceedings of the Sixth International Conference of InformationFusion,Cairns,Australia,2011:792-799.
[62] Clark D E,Bell J.Bayesian multiple target tracking inforw ard scan sonar image using thePHD filter[J].IEE Radar,Sonar and Navigation,2011,152(5):327-334.
[63] Panta K,Vo B-N,Singh S. Improved probability hypothesis density (PHD) filter formulti-target tracking[C].Proceedings of International Conference on Intelligent Sensing andInformation.Bangalore,India,2012:213-218.
[64]周刚.综放工作面喷雾降尘理论及工艺技术研究[D].泰安：山东科技大学，2009.
[65] Tan R T.Visibility in Bad Weather from a Single Image[C].Proc of the IEEE ComputerSociety Conference on Computer Vision and Pattern Recognition,Anchorage,USA,2008,:1-8.
[66] Fattal R.Single image dehazing [J]．ACM Transactions on Graphics,2008,27(3):721-729．
[67] Tarel J P,Hautière N.Fast Visibility Restoration from a Single Color or Gray Level Image [C].Proc of the12th IEEE International Conference on Computer Vision.Kyoto,Japan,2009:2201-2208.
[68] He Kaiming,Sun Jian,Tang Xiaoou.Single Image Haze Removal Using Dark Channel Prior
[C].Proc of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Miami,USA,2009:1956-1963.
[69]王多超，王永国，董雪梅.贝叶斯框架下的单幅图像去雾算法[J].计算机辅助设计与图形学学报，2010，22（10）：1756-1761.
[70]郭璠，蔡自兴.图像去雾算法清晰化效果客观评价方法[J].自动化学报，2012，38(009):1410-1419.
[71]Narasimhan S G,Nayar S K.Contrast Restoration of Weather Degraded Images.IEEETransaction on PatternAnalysis and Machine Intelligence,2006,25(6):713-724.
[72]耿则勋.自适应光学图像复原理论[M].北京：科学出版社，2010.
[73]刘卫华，白本督，赵小强.基于模糊相似度融合的图像复原算法[J].计算机辅助设计与图形学学报，2013，25（5）：616-621.
[74]陈香苹，李生红，苏波，等.基于图像噪声分析的计算机生成图像检测算法[J].光电子激光，2010，1（2）：293-297.
[75]胡学友.雾天降质图像的增强复原算法研究[D].合肥：安徽大学，2011.
[76]冯骢，达飞鹏，陈璋雯.一种改进的基于暗原色理论的去雾方法[J].东南大学学报，2012，42（9）：70-73.
[77]孙抗，汪渤，周志强.基于双边滤波的实时图像去雾技术研究[J].北京理工大学学报，2011，31（7）：810-813.
[78] LevinA,L ischinski D,WeissY.Aclosed form solution to natural image matting[C].Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Washington,USA:IEEE,2009:61-68.
[79]Oak Ley J P, Bu Hong. Correction of smiple contrast loss in color images[J]. ImageProcessing,2009,16(2):511-522.
[80]褚宏莉，李元祥，周则明，等.基于黑色通道的图像快速去雾优化算法[J].电子学报，2013，41（4）：791-796.
[81]Shuai Fang, Jiqing Zhan,Yang Cao,et al.Improved single image dehazing usingsegmentation[C].IEEE International Conference on Image Processing,Hong Kong,2010,3589-3592.
[82]孙伟，李大健，刘宏娟，等.基于大气散射模型的单幅图像快速去雾[J].光学精密工程，2013,21（4）：1040-1045.
[83]Elad M.On the Origin of the Bilateral Filter and Ways to Improve It[J].IEEE Trans on ImageProcessing,2005,11(10):1141-1151.
[84]Elad M.Retinex by Two Bilateral Filters[C]. Proc of the Scale-Space Conference.Hofgeismar,Germany,2007:217-229.
[85]陈龙，郭宝龙，毕娟，等.基于联合双边滤波的单幅图像去雾算法[J].北京邮电大学学报，2012,35（4）：19-22.
[86]易丽娅，鲁晓磊，王进军,等.图像复原的Bregman迭代双正则化方法[J].中国图象图形学报，2011,16（3）：350-355.
[87]Krishnan D,Tay T,Fergus R.Blind Deconvolution Using a Normalized SparsityMeasure[C].IEEE Conference on Computer Vision and Pattern Recognition.Colorado:IEEE,2011,233-240.
[88]稽晓强.图像快速去雾与清晰度恢复技术研究[D].长春：中国科学院长春光学精密机械与物理研究所，2012.
[89]王勇，谭义华，田金文.一种新的图像清晰度评价函数[J].武汉理工大学学报，2007,29（3）：124-126.
[90]陈勇，李愿，吕霞付等.视觉感知的彩色图像质量积极评价[J].光学精密工程，2013,21（3）：742-749.
[91]Zivkovic Z, Heijden F V D. Efficient adaptive density estimation per image pixel for the taskof background subtraction[J]. Pattern Recognition Letters,2008,27(7):773-780.
[92]张水发，丁欢，张文生.双模型背景建模与目标检测研究[J].计算机研究与发展，2011,48（11）：1983-1988.
[93] Stauffer C, Grimson W E L. Learning patterns of activity using real-time tracking[J]. IEEETransactions on PatternAnalysis and Machine Intelligence,2006,22(8):747757.
[94]仲元昌，蔡增增，赵贞贞，等.基于背景差法的几种背景建模方法的研究[J].计算机工程与应用，2012，48(24):62-66.
[95] Kristan M, Skocaj D, Leonardis A. Online kernel density estimation for interactiveI[leE9a6rE]nE iBn Tagrn[Jae]nr.sj aeImcet aiAgon,e sBan oudnr l IiVmniasa igPoe.n Computing,2011,28(7):11061116. PErfocceisesnint gp,a2r0ti1c0le,1fi9l(te9r)i:n2g48v0ia2sp4a9r0s.e kernel density estimation[J].
[97]张国权.基于视觉导航的智能车辆目标检测关键技术研究[D].兰州理工大学，2012.
[98]韦巍.基于广义Hough变换的目标跟踪算法研究[D].杭州：浙江大学，2013.
[99]王勇.基于统计方法的运动目标检测与跟踪技术研究[D].武汉：华中科技大学，2009.
[100]白亮.聚类学习的理论分析与高效算法研究[D].太原：山西大学，2012.
[101]Xu R,Wunsch D.Survey of clustering algorithms[J].IEEE Trancactions on NeuralNetworks,2008,16(3):645-678.
[102]熊子源，徐振海，张亮等.基于聚类算法的最优子阵划分方法研究[J].电子学报，2011,39（11）：2616-2620.
[103] WangWeina, Zhang Yunjie. On Fuzzy Cluster Validity Indices[J].Fuzzy Sets and Systems,2009,158(19):2095-2117.
[104]Kapp A V,Tibshirani R.Are clusters found in one dataset present in anotherdataset?[J].Biostatistics,2008,8(1):9-31.
[105]Zhou T,Zhang Y N,Yuan H J,et al.Rough K-means cluster with adaptiveparameters[C].Proceedings of the Sixth International Conference on Machine Learning andCybernetics.Piscataway,NJ,USA:IEEE,2009:3063-3068.
[106]闫爱云，李海朋，李晶皎等.视频运动目标提取的实现[J].东北大学学报，2011，2（11）：1558-1561.
[107]甘明刚，陈杰，刘劲等.一种基于三帧差分和边缘信息的运动目标检测方法[J],电子与信息学报，2010，2（4）.
[108]Huang Deng-yuan,Wang Chia-hung.Optimal mult-i level thresholding using a two stageOTSU optimization approach[J].Pattern Recognition Letters,2009,30:275-284.
[109] Sanina, Sanderson C, Lovell B C.Shadow detection: a survey and comparative evaluationof recent methods[J]. Pattern Recognition,2012,45(4):1684-1695.
[110]李广伦，殳伟群.运动目标的阴影实时检测和消除[J].小型微型计算机系统，2011，2（2）：361-364.
[111]王国良，梁德群，王演.基于区域与光照不变性的运动阴影检测算法[J].计算机应用，2007，7（9）：2152-2153.
[112]程佩青.数字信号处理教程[M].北京：清华大学出版社,2007.
[113]Tsaidm,Lin C T. Fastnormalized cross correlation for defect detection[J]. PatternRecognition Letters,2008,24:2625-2631.
[114]曹健，陈红倩，张凯等.结合区域颜色和纹理的运动阴影检测方法[J].机器人，2011,33f[（e1a19t5）u]：rSea6s[lv28-633.J]a.dCoorm pEu,tCear vVailslaioron aAnd,E Ibmraagheim Ui ndTe.Crsatsatn dsihnagd,2o0w08s,9e5g(m2e).ntation using invariant color
[116]姜明新.智能视频监控中目标跟踪技术研究[D].大连：大连理工大学，2013.
[117]闻帆.基于视觉的交通路口车辆智能检测技术研究[D].哈尔滨：哈尔滨工业大学，2010.
[118]曲昭伟.混合交通视频检测算法研究[D].长春：吉林大学，2009.
[119]Rao B,Zheng G,Chen T M,et al.An efficient hierarchical method for image shadowdetection[C].2nd International Workshop on Knowledge Discovery and Data Mining.Piscataway, NJ, USA: IEEE,2011:622-627.
[120]姚志均.目标跟踪系统中的鲁棒性研究[D].武汉：华中科技大学，2012.
[121] Avidan S.Support vector tracking[J].IEEE Transactions on PAMI,2007,26(8):1064-1072.
[122]郭烈，张明恒，李琳辉，等.一种基于支持向量机的行人识别方法研究[J].大连理工大学学报，2011,51（4）：604-609.
[123]贾静平，张飞舟，柴艳妹.Adaboost目标跟踪算法[J].模式识别与人工智能，2009,22（3）：475-480.
[124]文学志.基于机器学习的路面对象识别关键技术研究[D].沈阳：东北大学，2007.
[125]常甜甜.支持向量机学习算法若干问题的研究[D].西安：西安电子科技大学，2010.
[126]王爱平，万国伟，程志全.支持在线学习的增量式极端随机森林分类器[J].软件学报，211,22（9）：2059-2072.
[127]Galmeanu H,Andonie R.Implemetation Issues of an Incremental and DecrementalSVM[J].Lecture Notes in Computer Science,2011,5163:325-335.
[128]ZANDA M,BROWN G,FUMERA G, et al.Ensemble learning in linearly combinedclassifiers via negative correlation[C]MProceedings of MCS2007. Berlin,Germany:Springer-Verlag,2009:440O449.
[129]唐伟，周志华.基于Bagging的选择性聚类集成[J].软件学报，2005,16（4）：496-500.
[130]赵强利，蒋艳凰，徐明.选择性集成算法分类与比较[J].计算机工程与科学，2012,34（2）：134-138.
[131]张谢华，路梅，田敏.基于支持向量机的目标跟踪研究[J].计算机工程与设计，2011,32（12）：4210-4212.
[132]肖海军，王小非，洪帆等.基于特征选择和支持向量机的异常检测[J].华中科技大学学报，2008,36（3）：99-102.
[133]Wang X Y,Han T X,Yan S C.An HOG-LBP human detector with partial occlusionhandling[C].Proceedings of the3rdPacific Rim Symposium on Advances in Images and VideoTechnology.Heidelberg:Spinger,2010:37-47.
[134]Ning J F,Zhang L,Zhang David,et al.Robust object tracking using joint color-texturehistogram[J].International Journal of Pattern Recognition andArtificial Intelligence,2009,23(7):1245-1263.
[135]Kim P J,Chang H J,Choi J Y.Fast incremental learning for one-class support vectorclassifier using sample margin information[C].Proceedings of19thInternational Coference onPattern Recoginition.Tampa:IEEE,2011:1-4.
[136]秦传东.模糊与双重正则化支持向量机的研究及应用[D].西安：西安电子科技大学，2012.
[137]Wang Xiaodan,Wu Chongming,Bai Dongying,et al. A fast SVM incremental learningalgorithm based on the central convex hulls algorithm[C].Global Congress on intelligent Systems2009:IEEE Press,2009:472-475.
[138]张立，孟相如，马志强等.边界偏转覆盖增量支持向量机[J].北京邮电大学学报，2010,33（4）：30-33.
[139]Duangsoithong R,Windeatt T.Relevance and redundancy analysis for ensemble classifiers.In: Perner P, ed. Proc. of the Machine Learning and Data Mining in Pattern Recognition.Heidelberg: Springer-Verlag,2011:206-220.
[140]Mu T T,Nandi A K.Multiclass classification based on extended support vector datadescription[C].IEEE Transactions on System Man and Cybernetics-Part B:Cybernetics,2009,39(5):1206-1212.
[141]万琴.智能视觉监控中多运动目标检测与跟踪方法研究[D].长沙：湖南大学，2009.
[142]Wolf M T,Burdick J W.Multiple hypothesis tracking using clustered measurements[C].IEEEInternational Conference on Robotics andAutomation.Kobe:IEEE Press,2010:1321-1330.
[143]Charles K.卡尔曼滤波及其实时应用[M].北京：清华大学出版社，2013.
[144] Walter M R,Eustice R M,Leonard J J.Exactly Sparse Extended Information Filters forFeature-Based SLAM[J]. The International Journal of Robotics Research,2011,26(4):335-339.
[145]Kol S,Fossa B A,Scheic T S.Constrained nonlinear state estimation based on the UKFapproach[J]. Computer and Chemical Engineering,2009,(33):1386-1401.
[146]Rudolph der Merwe, Doucet A, De Freitas N, et al.The unscented particlefilter[A].Advances in Neural Information Proc Systems(NIPS13)[C]. http://speech.bme.ogi.edu/publications/ps/merwe00a.pdf,2010-03-10.
[147]刘也，余安喜，朱炬波等.加性噪声条件下的UKF算法[J].中国科学，2010,40（11）：1286-1296.
[148]Julier S J,Uhlmann J K. Unscented lteringand nonlinear estimation[J].Proceedings of theIEEE,2008,92(3):401-422.
[149]刘贵喜，马涛，陈石磊.应用最小偏度采样的UPF算法[J].光学精密工程，2008,16（4）：746-750.
[150]曹轶之.非高斯/非线性滤波算法研究及其在GPS动态定位中的应用[D].郑州：解放军信息工程大学，2012.
[151]Li H, Xu D, Jun J, et al. Sequence unscented Kalman filtering algorithm[C].The3rd IEEEConference on Industrial Electronics andApplications(ICIEA2010),Singapore,2008:1374-1378.
[152]Bowman C L,Morefield C L.Multisensor Fusion of Target Attributes andKinematics[C].Proceedings of the19thIEEE Conference on Decision and Control including theSymplsium onAdaptive Processes.San Diego:IEEE,2010:837-839.
[153]刘献如.视频图像序列目标跟踪算法及其应用研究[D].长沙：中南大学，2011.
[154]孔斌.快速连通域分析算法及其实现[J].模式识别与人工智能，2003,16（1）：110-114.
[155]Dunne P and Matuszewski B.Choice of similarity measure likelihood function andparameters for histogram based particle filter tracking in CCTV grey scale video[J]. Image andVision Computing,2011,29(2/3):178-189.
[156]李林.机动目标跟踪航迹生成的融合算法研究[D].长沙：国防科技大学，2008.
[157]汤军，孙伟.弹道目标跟踪的自适应多维分配相关算法[J].弹道学报：2011，23（2）：72-75.
[158]王智.机载多传感器数据融合技术研究[D].南京：南京理工大学，2010.