复杂气象条件下动态人群场景分析方法研究
|
摘要
基于计算机视觉的动态人群场景分析在智能视频监控、智能人群管理、公共空间设计、智能交通和智能环境、虚拟环境等领域有着巨大的应用潜力，特别是近年来大规模群体性公共安全事件频发对以智能视频监控为代表的人群行为分析提出了更高的要求。分析动态人群场景是分析以人群运动形式所表现出来的人群行为，即场景中所呈现出的人群运动状态。人群场景分析的内容在检测异常人群行为的基础上还应该包括跟踪人群运动状态的演进过程、分析运动状态变化规律和预报人群状态的未来发展趋势。目前通过视频进行人群运动状态分析面临的主要问题有两个：一是复杂天气造成的视频退化影响分析结果的准确性，二是缺乏利用计算机视觉理论与方法对人群运动状态的演变规律开展更为深入的研究。
本课题研究了复杂气象条件对提取人群运动速度和人群密度等主要特征的干扰现象，并从局部和整体两个角度研究了人群运动状态演化规律，提出了复杂气象视频分类与动态天气视频复原算法，然后利用增强后的视频数据建立人群流量方程和人群运动状态无向图两种人群运动分析模型，解决了雨雪天气视频退化现象对人群运动状态分析的影响和如何对未来人群运动状态进行预测的问题。研究工作的主要贡献体现在以下四个方面：
一、提出一种基于视频序列短时相关性和平均相关性特征的分级天气类型分类算法。在像素亮度值的时间序列上建立自相关函数提取场景特征，构造线性分类器并使用CART决策树将室外场景分类为静态、动态和不稳定三类天气。分类结果为下面的视频复原方法提供天气视频分类前提，即后续的去雨雪方法是基于正确的动态天气视频分类之上。
二、提出一种滑动窗口序列在线PCA投影降噪和二次目标分割的视频去雨雪(动态天气)方法。室外监控视频常常受到各种复杂气象干扰（如雨、雪、雾、光照变化和低温等），造成监控视频发生退化，影响了场景分析的准确性。有必要在抑制天气干扰算法方面展开研究，并把它纳入群体系统模型中。针对动态天气类型分析室外视频的噪声特点、前景目标运动性质的差异，提出一种窗口序列在线PCA投影降噪的二次分割视频去雨雪方法。将具有不同运动性质的移动目标（如人、车、镜头前的雨滴和雪花等）精确分类，定位雨滴和雪花区域后使用对应的背景图像修饰来消除不利天气对视频的影响，为后续的人群运动分析提供增强后的视频数据。
三、为避免复杂的个体目标分割和跟踪运算，提出一种无需前景分割的局部人群状态分析方法。把视频局部区域在时-空上视为一个独立的线性动态系统，通过混合动态纹理方法在空间上对局部人群分类得到人群密度属性；使用主路径跟从方法计算局部人群运动速度，然后依据人群密度和运动速度属性建立人群流量方程，用该方程可分析局部区域内人群密度、速度和流量之间的变化关系。该模型定量描述了局部人群运动状态、克服了群体运动的不确定性，可用于人群场景中重点监视区域的异常事件检测。
四、针对用于人群场景分析的参数模型缺乏对未来人群运动状态的预测手段和泛化能力不强等问题，提出一种非参数聚类的人群异常状态检测与预测的图分析方法。首先，计算场景的速度场作为基本特征数据，使用改进的均值飘移(Mean Shift)聚类算法将速度特征聚类，得到不同运动类别的聚类中心。然后以聚类中心为顶点、各中心间距离(速度差)为边构成无向图，在时间域分析图边权重对称矩阵的变化，比较观测值和预测值之间的背离程度可以检测当前和预测未来一段时间内人群运动状态。由于采用非参数聚类，避免了更换人群场景时重新设置初值、重新挑选训练样本和参数估计等繁杂工作，克服了参数估计精度问题产生的检测和预报误差。
前两项内容属于视频的预处理过程，完成天气分类并针对分类结果中的动态天气进行视频复原工作，为后续的人群行为分析提供较高质量的视频数据；后两个研究内容为动态人群场景建模过程，分别采用局部和整体方法描述了人群运动状态。
本课题在人群运动场景分析方面的研究工作，较好地解决了受雨雪气象条件干扰的视频复原问题，为模型提供增强的输入视频；提出人群流量方程分析局部人群运动状态，实现了重点区域的差别化监控；非参数化的运动特征分类方法和图分析算法增强了人群运动状态分析方法的泛化能力，为基于计算机视觉的人群行为分析提供了一种在整体上监测和预报人群运动状态的方法。
The analysis of dynamic crowd scene based on computer vision can be appliedbroadly to several of fields including intelligent surveillance, crowd management,public space design, intelligent transportation, virtual environment, etc. Thefrequent occurrences of large-scale mass incidents come to challenges tointelligent surveillance, in recent years, based on crowd motion analysis. Priormodels concerning crowd abnormity detection does not satisfy the public safesituation, more researches, such as tracking the population status andevolutionary process, analysis of variation of crowd motion, predicting futuretrends and so on, are required. Modeling crowd motion scene facing two majorchallenges, one is the negative impact on analysis result caused by visiondegradation of complicated weather conditions, and the other is the lack ofin-depth investigation for the law of crowd motion based on computer vision.
In the thesis, the author discusses how the bad weathers interfere featuresextraction of crowd scene in videos and investigates the law of crowd motion inways of local and holistic aspects; and proposes a classification method ofweather types and a video (captured in dynamic weather environment) restorationmeans. Furthermore, performs a comprehensive analysis for crowd motion statesusing enhanced video dataset. Finally, the problem of video degradation byrain/snow is solved and a crowd scene model is built by using a crowd flowequation and graphic analysis. The proposed model can be used to detect andpredict feature crowd status.
Aiming at building a complete crowd motion model in complicated weatherconditions, the main contributions of the thesis is reflected in the following fouraspects:
1. A classification method of outdoor videos in complex weather conditionsbased on short time correlation and mean of video is proposed. The studyformulates an autocorrelation function in the time domain to extract pixelintensity feature, and classifying videos employing a linear classifier with aclassification and regression-tree (CART). As a result, the input video isclassified to three types: static, dynamic and non-stationary weather according to different weather conditions. It is worth noting that the dynamic weather video isused as input data in the following study.
2. A video denoising and rain/snow removal method is proposed by a videowindow series on-line PCA projection algorithm on video window series.Outdoor video often suffer interferences from bad weathers (rain, snow, fog,illumination changing, high and low temperature, etc.), the degraded videosreduce the accuracy of the crowd scene analysis. Therefore, it is necessary todevelop algorithm about inhibition of weather disturbances as a part of wholecrowd model. Using the proposed model, the moving objects with differentmotion properties can be separated and those non-interesting regions are replacedby background. Consequently, the model outputs enhanced video for subsequentapplication of crowd motion analysis.
3. To avoid complex calculation of object segmentation and tracking, a localanalysis method of crowd status without foreground segmentation is proposed.Each local area in a video is treated as an independent linear dynamic system(LDS) in temporal-spatial domain. Local crowd groups are classified employingthe mixture of dynamic texture to get crowd density, and employing a main pathfollowing method to obtain local crowd speed. Next, using the density and speedproperties a crowd flow equation, which holds the relations between crowddensity, speed and flow, is built. The equation overcomes the drawbacks of theuncertainty of crowd motion, describes the local scene state and can be used todetecting abnormal events in key areas.
4. A non-parameter clustering graph analysis is put forward in crowdabnormity detecting and motion predicting to fit the lack of generalizationcapability in most existing crowd models. The velocity fields are extracted as abasic dataset in which cluster centers are obtained via Mean Shift (MS) clusteringalgorithm, and all centers and the Euclidean distances between them form anundirected graph that reveals the insight of crowd motion status. Thus, theabnormal events could be detected and future states of the crowd could bepredicted by computing distributions of all graph vertexes in feature space, aswell as testing the deviations between observations and estimates of the edgeweights matrix dynamic system. The proposed method need not reinitialize model,retrain new samples and estimate parameters.
The first two items are video preprocessing that completes the weather videoclassification and dynamic weather video restoration task, their output providesmore quality video data for subsequent crowd behavior analysis. The latter twoitems involve crowd scene modeling by local and holistic approaches.
The research involves video restoration, local crowd state analysis andpredicting future crowd behavior via a graph method. All parts form an integralmodel of dynamic crowd scene that has the capacity of video denoising, imageenhancement, crowd abnormity detection and state forecasting. It provides afeasible monitoring and forecasting way for crowd behaviors analysis based oncomputer vision techniques.
本课题研究了复杂气象条件对提取人群运动速度和人群密度等主要特征的干扰现象，并从局部和整体两个角度研究了人群运动状态演化规律，提出了复杂气象视频分类与动态天气视频复原算法，然后利用增强后的视频数据建立人群流量方程和人群运动状态无向图两种人群运动分析模型，解决了雨雪天气视频退化现象对人群运动状态分析的影响和如何对未来人群运动状态进行预测的问题。研究工作的主要贡献体现在以下四个方面：
一、提出一种基于视频序列短时相关性和平均相关性特征的分级天气类型分类算法。在像素亮度值的时间序列上建立自相关函数提取场景特征，构造线性分类器并使用CART决策树将室外场景分类为静态、动态和不稳定三类天气。分类结果为下面的视频复原方法提供天气视频分类前提，即后续的去雨雪方法是基于正确的动态天气视频分类之上。
二、提出一种滑动窗口序列在线PCA投影降噪和二次目标分割的视频去雨雪(动态天气)方法。室外监控视频常常受到各种复杂气象干扰（如雨、雪、雾、光照变化和低温等），造成监控视频发生退化，影响了场景分析的准确性。有必要在抑制天气干扰算法方面展开研究，并把它纳入群体系统模型中。针对动态天气类型分析室外视频的噪声特点、前景目标运动性质的差异，提出一种窗口序列在线PCA投影降噪的二次分割视频去雨雪方法。将具有不同运动性质的移动目标（如人、车、镜头前的雨滴和雪花等）精确分类，定位雨滴和雪花区域后使用对应的背景图像修饰来消除不利天气对视频的影响，为后续的人群运动分析提供增强后的视频数据。
三、为避免复杂的个体目标分割和跟踪运算，提出一种无需前景分割的局部人群状态分析方法。把视频局部区域在时-空上视为一个独立的线性动态系统，通过混合动态纹理方法在空间上对局部人群分类得到人群密度属性；使用主路径跟从方法计算局部人群运动速度，然后依据人群密度和运动速度属性建立人群流量方程，用该方程可分析局部区域内人群密度、速度和流量之间的变化关系。该模型定量描述了局部人群运动状态、克服了群体运动的不确定性，可用于人群场景中重点监视区域的异常事件检测。
四、针对用于人群场景分析的参数模型缺乏对未来人群运动状态的预测手段和泛化能力不强等问题，提出一种非参数聚类的人群异常状态检测与预测的图分析方法。首先，计算场景的速度场作为基本特征数据，使用改进的均值飘移(Mean Shift)聚类算法将速度特征聚类，得到不同运动类别的聚类中心。然后以聚类中心为顶点、各中心间距离(速度差)为边构成无向图，在时间域分析图边权重对称矩阵的变化，比较观测值和预测值之间的背离程度可以检测当前和预测未来一段时间内人群运动状态。由于采用非参数聚类，避免了更换人群场景时重新设置初值、重新挑选训练样本和参数估计等繁杂工作，克服了参数估计精度问题产生的检测和预报误差。
前两项内容属于视频的预处理过程，完成天气分类并针对分类结果中的动态天气进行视频复原工作，为后续的人群行为分析提供较高质量的视频数据；后两个研究内容为动态人群场景建模过程，分别采用局部和整体方法描述了人群运动状态。
本课题在人群运动场景分析方面的研究工作，较好地解决了受雨雪气象条件干扰的视频复原问题，为模型提供增强的输入视频；提出人群流量方程分析局部人群运动状态，实现了重点区域的差别化监控；非参数化的运动特征分类方法和图分析算法增强了人群运动状态分析方法的泛化能力，为基于计算机视觉的人群行为分析提供了一种在整体上监测和预报人群运动状态的方法。
The analysis of dynamic crowd scene based on computer vision can be appliedbroadly to several of fields including intelligent surveillance, crowd management,public space design, intelligent transportation, virtual environment, etc. Thefrequent occurrences of large-scale mass incidents come to challenges tointelligent surveillance, in recent years, based on crowd motion analysis. Priormodels concerning crowd abnormity detection does not satisfy the public safesituation, more researches, such as tracking the population status andevolutionary process, analysis of variation of crowd motion, predicting futuretrends and so on, are required. Modeling crowd motion scene facing two majorchallenges, one is the negative impact on analysis result caused by visiondegradation of complicated weather conditions, and the other is the lack ofin-depth investigation for the law of crowd motion based on computer vision.
In the thesis, the author discusses how the bad weathers interfere featuresextraction of crowd scene in videos and investigates the law of crowd motion inways of local and holistic aspects; and proposes a classification method ofweather types and a video (captured in dynamic weather environment) restorationmeans. Furthermore, performs a comprehensive analysis for crowd motion statesusing enhanced video dataset. Finally, the problem of video degradation byrain/snow is solved and a crowd scene model is built by using a crowd flowequation and graphic analysis. The proposed model can be used to detect andpredict feature crowd status.
Aiming at building a complete crowd motion model in complicated weatherconditions, the main contributions of the thesis is reflected in the following fouraspects:
1. A classification method of outdoor videos in complex weather conditionsbased on short time correlation and mean of video is proposed. The studyformulates an autocorrelation function in the time domain to extract pixelintensity feature, and classifying videos employing a linear classifier with aclassification and regression-tree (CART). As a result, the input video isclassified to three types: static, dynamic and non-stationary weather according to different weather conditions. It is worth noting that the dynamic weather video isused as input data in the following study.
2. A video denoising and rain/snow removal method is proposed by a videowindow series on-line PCA projection algorithm on video window series.Outdoor video often suffer interferences from bad weathers (rain, snow, fog,illumination changing, high and low temperature, etc.), the degraded videosreduce the accuracy of the crowd scene analysis. Therefore, it is necessary todevelop algorithm about inhibition of weather disturbances as a part of wholecrowd model. Using the proposed model, the moving objects with differentmotion properties can be separated and those non-interesting regions are replacedby background. Consequently, the model outputs enhanced video for subsequentapplication of crowd motion analysis.
3. To avoid complex calculation of object segmentation and tracking, a localanalysis method of crowd status without foreground segmentation is proposed.Each local area in a video is treated as an independent linear dynamic system(LDS) in temporal-spatial domain. Local crowd groups are classified employingthe mixture of dynamic texture to get crowd density, and employing a main pathfollowing method to obtain local crowd speed. Next, using the density and speedproperties a crowd flow equation, which holds the relations between crowddensity, speed and flow, is built. The equation overcomes the drawbacks of theuncertainty of crowd motion, describes the local scene state and can be used todetecting abnormal events in key areas.
4. A non-parameter clustering graph analysis is put forward in crowdabnormity detecting and motion predicting to fit the lack of generalizationcapability in most existing crowd models. The velocity fields are extracted as abasic dataset in which cluster centers are obtained via Mean Shift (MS) clusteringalgorithm, and all centers and the Euclidean distances between them form anundirected graph that reveals the insight of crowd motion status. Thus, theabnormal events could be detected and future states of the crowd could bepredicted by computing distributions of all graph vertexes in feature space, aswell as testing the deviations between observations and estimates of the edgeweights matrix dynamic system. The proposed method need not reinitialize model,retrain new samples and estimate parameters.
The first two items are video preprocessing that completes the weather videoclassification and dynamic weather video restoration task, their output providesmore quality video data for subsequent crowd behavior analysis. The latter twoitems involve crowd scene modeling by local and holistic approaches.
The research involves video restoration, local crowd state analysis andpredicting future crowd behavior via a graph method. All parts form an integralmodel of dynamic crowd scene that has the capacity of video denoising, imageenhancement, crowd abnormity detection and state forecasting. It provides afeasible monitoring and forecasting way for crowd behaviors analysis based oncomputer vision techniques.
引文
[1] B. Zhan, D N. Monekosso, P. Remagnino, et al. Crowd Analysis: ASurvey[J]. Machine Vision and Applications.2008,18(5-6):345~357
[2] D. Helbing, L. Buzna, A. Johansson, et al. Self-organized Pedestrian CrowdDynamics: Experiments, Simulations, and Design Solutions[J].Transportation science.2005,39(1):1~24
[3] A.N. Marana, S.A. Velastin, L.F. Costa, R.A. Lotufo. Automatic estimationof crowd density using texture[J]. Safety Science.1998,28(3):165~175
[4] N. Dong, F. Liu, Z. Li. Crowd density estimation using sparse texturefeatures[J]. Journal of Convergence Information Technology.2010,5(6):125~137
[5] W. Li, X. Wu, K. Matsumoto, H.A. Zhao. Crowd density estimation: Animproved approach[C]. IEEE10th International Conference on SignalProcessing (ICSP’10).2010:1213~1216
[6] X. Wu, G. Liang, K.K. Lee, Y. Xu. Crowd density estimation using textureanalysis and learning[C]. IEEE International Conference on Robotics andBiomimetics(ROBIO'06).2006:214~219
[7] V. Rabaud, S. Belongie. Counting crowded moving objects[C]. ComputerVision and Pattern Recognition(CVPR’06), IEEE Computer SocietyConference on.17-22June2006, New York, NY, USA.2006,1:705~711
[8] H. Song, L. Xin, Y.Z. Chen, Y.Y. Gao. Weather identifying system basedon vehicle video image[C]. Intelligent Control and Automation (WCICA),9th World Congress on.21-25June2011, Taipei TaiWan.2011:172~175
[9] M. Roser, F. Moosmann. Classification of weather situations on single colorimages[C]. Intelligent Vehicles Symposium,2008IEEE. Eindhoven,Netherlands, June4-6,2008:798~803
[10] http://www.lyyn.com/showroom/applications/fog-and-haze/. LYYN,2012
[11] B.K.P. Horn, B.G. Schunck. Determining optical flow[J]. Artificialintelligence.1981,17(1-3):185~203
[12] A.B. Chan, Z.S.J. Liang, N. Vasconcelos. Privacy preserving crowdmonitoring: Counting people without people models or tracking[C]. IEEEConference on Computer Vision and Pattern Recognition(CVPR’08).2008:1~7
[13] S.F. Lin, J.Y. Chen, H.X. Chao. Estimation of number of people in crowdedscenes using perspective transformation[J]. Systems, Man and Cybernetics,Part A: Systems and Humans, IEEE Transactions on.2001,31(6):645~654
[14] V. Zlokolica, A. Pizurica, W. Philips. Video denoising using multiple classaveraging with multiresolution[J]. Visual Content Processing andRepresentation.2003,2849:172~179
[15] I.W. Selesnick, K.Y. Li. Video denoising using2D and3D dual-treecomplex wavelet transforms[J]. Wavelets: Applications in Signal and ImageProcessing X.2003,5207:607~618
[16] H.J. Seo, P. Milanfar. Video denoising using higher order optimalspace-time adaptation[C]. Acoustics, Speech and Signal Processing,2008.ICASSP2008. Santa Cruz, CA, March312008-April42008,1249~1252
[17] H. Ji, C. Liu, Z. Shen, Y. Xu. Robust video denoising using low rank matrixcompletion[C]. Computer Vision and Pattern Recognition (CVPR’10), IEEEConference on. June13-18,2010, San Francisco, CA, USA:1791~1798
[18] J. Chen, C.K. Tang. Spatio-temporal Markova Random Field for videodenoising[C]. In: Proc. of the IEEE Conference on Computer Vision andPattern Recognition(CVPR'07).18-23June2007, Minneapolis, Minnesota,USA.1~8
[19] F. Luisier, T. Blu, M. Unser. SURE-LET for orthonormal wavelet-domainvideo denoising[J]. IEEE Trans. Circuits and Systems for Video Technology.2010,20(6):913~919
[20] G. Varghese, Z. Wang. Video denoising based on a spatiotemporal Gaussianscale mixture model[J]. Circuits and Systems for Video Technology, IEEETransactions on.2010,20(7):1032~1040
[21] R. Eslami, X. L Wu. Video denoising using3-D Hybrid Wavelets andDirectional filter banks[C]. Image Processing2008(ICIP2008), San Diego,October12-15,2008,2340~2343
[22] C. Liu, W. Freeman. A high-quality video denoising algorithm based onreliable motion estimation[J]. Computer Visio (ECCV’10).2010:706~719
[23] V. Zlokolica, W. Philips. Motion and Detail Adaptive Denoising in Video[J].Proc. SPIE—Image Processing: Algorithms and Systems III,2004(5298):403~412
[24] H. Y. Cheong, A. Tourapis, J. Llach, J. Boyce. Adaptive spatio-temporalfiltering for video-denoising[J]. IEEE Trans. Image Processing.2004,2:965~968
[25] J. Boulanger, C. Kervrann, P. Bouthemy. Space–time adaptation forpatch-based image sequence restoration[J]. IEEE Trans. PAMI,2007(29):1096~1102
[26] S. Kindermann, S. Osher, P.W. Jones. Deblurring and denoising of imagesby nonlocal functionals[J]. Multiscale Modeling Simulation,2005(4):1091~1115
[27] A. Elmoataz, O. Lezoray, S. Bougleux. Nonlocal discrete regularization onweighted graphs: a framework for image and manifold processing[J]. IEEETrans. Image Process.2008,17(7):1047~1060
[28] M. Ghoniem, Y. Chahir, A. Elmoataz. Video denoising and simplificationvia discrete regularization on graph[C]. Advanced Concepts for IntelligentVision Systems (ACIVS’08),2008:380~389
[29] M. Ghoniem, Y. Chahir, A. Elmoataz. Video denoising via discreteregularization on graphs[C]. ICPR’08, Florida, USA, December8-11,2008:1~4
[30] M. Protter, M. Elad, H. Takeda, P. Milanfar, Generalizing the nonlocal-means to super-resolution reconstruction[J]. IEEE Trans. Image Process.2009,18(1):36~51
[31] A. Buades, B. Coll, J. M. Morel. Nonlocal image and movie denoising[J].Comput Vision,2008,76(2):123~139
[32] M. Ghoniem, Y. Chahir, A. Elmoataz. Nonlocal video denoising,simplification and inpainting using discrete regularization on graphs[J].Signal Processing.2008,90(8):2445~2455
[33] R.S. Lee, C. Craun, M. Robinson, M. Phaneuf. Evaluating the impacts ofhaze on air traffic operations[C]. The92nd American MeteorologicalSociety Annual Meeting. January22-26,2012, New Orleans, LA, USA.http://ams.confex.com/ams/92Annual/webprogram/Paper201227.html(2012)
[34] K.M. He, J. Sun, X.O Tang. Single Image Haze Removal Using DarkChannel Prior[J]. IEEE Transactions on Pattern Analysis and MachineIntelligence.2011,33(12):2043~2051
[35] S. Shwartz, E. Namer, Y.Y. Schechner. Blind Haze Separation[C]. Proc.IEEE Conf. Computer Vision and Pattern Recognition.17-22June2006,New York, NY, USA.2006,2:1984~1991
[36] X. Zhang, H. Li, Y. Qi, W.K. Leow, T.K. Ng. Rain removal in video bycombining temporal and chromatic properties[C]. Proceedings of the2006IEEE International Conference on Multimedia and Expo. Washington DC:IEEE Computer Society,2006:461~464
[37] P. Barnum, T. Kanade, S.G. Narasimhan. Spatio-temporal frequencyanalysis for removing rain and snow from videos [C]//Proceedings of theFirst International Workshop on Photometric Analysis for Computer Vision,in conjunction with International Conference of Computer Vision(ICCV2007). Rio de Janeiro: INRIA,2007:8p
[38] P. Barnum, Narasimhan S, Kanade T. Analysis of rain and snow infrequency space [J]. International Journal of Computer Vision.2009,86(2-3):256~274
[39] X. S. Kong, A. D. Zhang, Y. G. Qian. Haze and cloud cover recognition andremoval for serial Landsat images. Proc. SPIE8006,80061K (2011);doi:10.1117/12.902875(Remote Sensing Image Processing, GeographicInformation Systems, and Other Applications)
[40] F. Cogun, A.E. Cetin. Object tracking under illumination variations using2D-cepstrum characteristics of the target[C]. Multimedia Signal Processing(MMSP),2010IEEE International Workshop on.4-6Oct.2010, Saint-Malo,France.2010:521~526
[41] Freedman D, Turek M W. Illumination-invariant tracking via graph cuts [C]//Proceedings of the2005IEEE Computer Society Conference on ComputerVision and Pattern Recognition. Washington DC: IEEE Computer Society,2005(2):10~17
[42] D. Zang. Illumination invariant object tracking based on multiscalephase[C]. Image Processing (ICIP),201017th IEEE InternationalConference on.26-29Sept.2010, Shanghai, China.2010:357~360
[43] M. Haindl, P. Vacha. Illumination Invariant Texture Retrieval[C]. PatternRecognition(ICPR2006),18th International Conference on.20-24Aug.2006, Hong Kong, China.2006,3:276~279
[44]单勇，王润生.适应灰度和光照变化的运动目标跟踪方法[J].计算机辅助设计与图形学学报,2006,18(2):283~288
[45] Slater D, Healey G. The illumination-invariant recognition of3D objectsusing local color invariants [J]. IEEE Transactions on Pattern Analysis andMachine Intelligence,1996,18(2):206~210
[46] T. Hou, S. Wang, H. Qin. Vehicle matching and recognition under largevariations of pose and illumination[C]. Computer Vision and PatternRecognition Workshops,2009. CVPR Workshops2009. IEEE ComputerSociety Conference on.20-25June2009, Miami, FL, USA.2009:24~29
[47] P. Felzenszwalb, R. Girshick, D. McAllester, D. Ramanan. Object detectionwith discriminatively trained part based models[J]. IEEE Trans. PAMI.2010,32(9):1627~1645
[48] R. Mehran, A. Oyama, M. Shah. Abnormal crowd behavior detection usingsocial force model[C]. Computer Vision and Pattern Recognition(CVPR’09)IEEE Conference on.20-25June2009, Miami, FL.2009:935~942
[49] M. Rodriguez, I. Laptev, J. Sivic, J.Y. Audibert. Density-aware persondetection and tracking in crowds[C]. Computer Vision (ICCV’11), IEEEInternational Conference on.6-13Nov.2011, Barcelona, Spain.2011:2423~2430
[50] O. Tuzel, F. Porikli, P. Meer. Pedestrian Detection via Classification onRiemannian Manifolds[J]. IEEE Trans. PAMI.2008,30(10):1713~1727
[51] S. Wang, Z.J. Miao. Anomaly detection in crowd scene using historicalinformation[C]. Intelligent Signal Processing and Communication Systems(ISPACS), International Symposium on.2010:1~4
[52] S. Stalder, H. Grabner, L.V. Gool. Exploring context to learn scene specificobject detectors[C]. Proceedings11th IEEE International Workshop onPETS. Miami, June25,2009:63~70
[53] L. Kratz, K. Nishino. Anomaly detection in extremely crowded scenes usingspatio-temporal motion pattern models[C]. Computer Vision and PatternRecognition(CVPR’09). IEEE Conference on,2009:1446~1453
[54] K.M. Zeitz, H.M. Tan, M. Grief, P.C. Couns, C.J. Zeitz. Crowd behavior atmass gatherings: a literature review[J]. Prehospital and disaster medicine.2010,24(1):32~38
[55] M. Mancas. Attention-Based Dense Cropwds Analysis[C]. Image Analysisfor Multimedia Interactive Services (WIAMIS), the11th InternationalWorkshop on.2010:1~4
[56] D. Ryan, S. Denman, C. Fookes, S. Sridharan. Crowd counting usingmultiple local features[J]. Digital Image Computing: Techniques andApplications.2009,81~88
[57] S. Denman, V. Chandran, and S. Sridharan. An adaptive optical flowtechnique for person tracking systems[J]. Elsivier Pattern RecognitionLetters.2007,28(10):1232~1239
[58] W. Ge R. Collins, R. Ruback. Vision-based analysis of small groups inpedestrian crowds[J]. IEEE Trons. PAMI.2011,99:1~14
[59] M.C. Chang, N. Krahnstoever, S. Lim, Y.T. Group Level ActivityRecognition in Crowded Environments across Multiple Cameras[C].Proceedings of the7th IEEE International Conference on Advanced Videoand Signal Based Surveillance (AVSS’10),29August-1September2010,Boston, Massachusetts.2010:56~63
[60] S. Ali, M. Shah. A Lagrangian particle dynamics approach for crowd flowsegmentation and stability analysis[C]. Computer Vision and PatternRecognition(CVPR'07). IEEE Conference on.2007:1~6
[61] A.B. Chan, N. Vasconcelos. Modeling, clustering, and segmenting videowith mixtures of dynamic textures[J]. IEEE Trons. PAMI.2008,30(5):909~926
[62] E.L. Andrade, R.B. Fisher. Simulation of crowd problems for computervision[C]. First International Workshop on Crowd Simulation.2005,3:71~80
[63] N. Pelechano, J.M. Allbeck, N.I. Badler. Controlling individual agents inhigh-density crowd simulation[C]. Proceedings of the2007ACMSIGGRAPH/Eurographics symposium on Computer animation.2007:99~108
[64] C. Peters, C. Ennis. Modeling groups of plausible virtual pedestrians[J].Computer Graphics and Applications.2009,29(4):54~63
[65] R A. Skov. Impacts of haze on high-capacity airport operations. The14thConference on Aviation, Range, and Erospace Meteorology. NOAA/NWS,Hampton, GA,2010
[66] Medina, Juan C, Chitturi, Madhav, Benekohal, Rahim F. Effects of fog,snow, and rain on video detection systems at intersections. TransportationLetters, International Journal of Transportation Research.2010,2(1):1~12
[67] X.L. Yu, C. B Xiao, M. Deng, P. Li. A Classification Algorithm toDistinguish Image as Haze or Non-haze[C]. In Proceedings of the SixthInternational Conference on Image and Graphics (ICIG’11). Hefei, China.August12-15,2011:286~289
[68] A. Elgammal, D. Harwood, L. Davis. Non-parametric model for backgroundsubtraction[C]. Computer Vision(ECCV2000, Springer). Dublin, Ireland,June26-July1,2000:751~767
[69] M.R. Blanton, E. Kazin, D. Muna, B.A. Weaver, A. Price-Whelan.Improved background subtraction for the Sloan Digital Sky Surveyimages[J]. The Astronomical Journal.2011,142:31pages, IOP Publishing.doi:10.1088/0004-6256/142/1/31
[70] Y. Benezeth, P.M. Jodoin, B. Emile, H. Laurent, C. Rosenberger.Comparative study of background subtraction algorithms[J]. Journal ofElectronic Imaging.2010,19:033003
[71] A. Lipton, H. Fujiyoshi, R. Patil. Moving target classification and trackingfrom real-time video[C]. In: Proc IEEE Workshop on Applications ofComputer Vision, Princeton, NJ,1998:8~14
[72] á. Bayona, J. SanMiguel, J. Martinez. Stationary foreground detectionusing background subtraction and temporal difference in videosurveillance[C]. Image Processing (ICIP’10), the17th IEEE InternationalConference on. September26-292010, Hong Kong, China,2010:4657~4660
[73] J.L. Barron, D.J. Fleet, S.S. Beauchemin. Performance of optical flowtechniques[J]. International journal of computer vision.1994,12(1):43~77
[74] S. Baker, D. Scharstein, J.P. Lewis, S. Roth,M.J. Black, R. Szeliski. Adatabase and evaluation methodology for optical flow[J]. InternationalJournal of Computer Vision.2011,92(1):1~31
[75] M.B. Holte, T.B. Moeslund, P. Fihl. View-invariant gesture recognitionusing3D optical flow and harmonic motion context[J]. Computer Visionand Image Understanding.2010,114:1353~1361
[76] K.P. Murphy. Dynamic Bayesian networks[J]. Probabilistic GraphicalModels, in M. Jordan.2002. http://www.cs.ubc.ca/~murphyk/Papers/dbn-chapter.pdf (2012)
[77] Z. Zeng, Q. Ji. Knowledge based activity recognition with dynamicBayesian network[J]. Computer Vision.2010:532~546
[78] H.I. Suk, B.K. Sin, S.W. Lee. Hand gesture recognition based on dynamicBayesian network framework[J]. Pattern Recognition,2010,43(9):3059~3072
[79] J. Cortes. Distributed Kriged Kalman Filter for SpatialEstimation[J]. Automatic Control, IEEE Transactions on.2009,12:2816~2827
[80] M. Mehta, C. Goyal, M.C. Srivastava, R.C Jain. Real time object detectionand tracking: Histogram matching and Kalman filter approach[C]. Computerand Automation Engineering (ICCAE’10), The2nd InternationalConference on.26-28Feb.2010, Singapore.2010,5:796~801
[81] H. Medeiros, J. Park, A. Kak. Distributed Object Tracking Using aCluster-Based Kalman Filter in Wireless Camera Networks[J]. IEEE Journalof Selected Topics in Signal Processing.2008,2(4):448~463
[82] Y.M. Lui, J.R. Beveridge, L.D. Whitley. A novel appearance model andadaptive condensation algorithm for human face tracking[C]. Biometrics:Theory, Applications and Systems(BTAS’08), the2nd IEEE InternationalConference on. Sept.29-Oct.1,2008, Arlington, VA:1~7
[83] L. Dan, O. Jun. Hand-gesture extraction and recognition from the videosequence acquired by a dynamic camera using condensation algorithm[C].Proc. SPIE7252,72520S (2009);19January2009, San Jose, CA, USA.http://dx.doi.org/10.1117/12.806108(2012)
[84] N. Hauti`ere, R. Labayrade, D. Aubert. Real-time disparity contrastcombination for onboard estimation of the visibility distance[J]. IEEETransactions on Intelligent Transportation Systems.2006(7):201~212
[85] E. Dumont, V. Cavallo. Extended photometric model of fog effects on roadvision[J]. Transp. Transportation Research Record: Journal of theTransportation Research Board.2004(1862):77~81
[86] H. Koschmieder. Theorie der Horizontalen Sichtweite. Beitr. Phys. FreienAtm.1924,12:171~181
[87] Narasimhan S G, Nayar S K. Vision and the Atmosphere[J]. InternationalJournal of Computer Vision.2002,48(3):233~254
[88] K. Miller. Least squares method for ILL-posed problems with a prescribedbound[J]. International Journal of Math Analysis.1990,(1):52~74
[89] M. Egmont-Petersen, D. de Ridder, H. Handels. Image processing withneural networks--a review[J]. J. of Pattern Recognition.2002,(35):2279~2301
[90] S.Z. Li. Markov Random Field Modeling in Image Analysis(ThirdEdition)[M]. Springer-Verlag New York Inc.2009.371p
[91] R. Rajagopalan. Synthesizing processed video by filtering temporalrelationships. IEEE Trans. on Image Processing.2002,11(1):26~36
[92] M. Mahmoudi, G. Sapiro. Fast image and video denoising via nonlocalmeans of similar neighborhoods[J]. Signal Processing Letters, IEEE.2005,12(12):839~842
[93] H. Faraji, M. J. MacLean. CCD noise removal in digital image[J]. IEEETransactions on Image Processing.2006,15(9):2676~2685
[94] K. Garg, S.K. Nayar. Vision and rain [J]. International Journal of ComputerVision,2007,75(1):3~27
[95] K. Garg, S. K. Nayar. Detection and removal of rain from videos[C]//Proceedings of the2004IEEE Computer Society Conference on ComputerVision and Pattern Recognition. Washington DC: IEEE Computer Society,2004(1):528~535
[96] K. Garg, S. K. Nayar. Photorealistic rendering of rain streaks[J]. ACMTransactions on Graphics,2006,25(3):996~1002
[97] Brewer N, Liu N. Using the Shape Characteristics of Rain to Identify andRemove Rain from Video[J]. Structural, Syntactic, and Statistical PatternRecognition.2008,5342:451~458
[98] G. B. Foote, P. S. Du Toit. Terminal Velocity of Raindrops Aloft[J]. Journalof Applied Meteorology.1969,8(2):249~253
[99] B.K.P. Horn, B.G. Schunck. Determining optical flow[J]. Artificialintelligence.1981,17(1-3):185~203
[100] A.B. Chan, N. Vasconcelos. Mixtures of dynamic textures[C].Computer Vision(ICCV’05).2005,1:641~647
[101]V. Lempitsky, A. Zisserman. Learning To Count Objects in Images[J]. InNIPS2010:1~9
[102]O. Masoud, N.P. Papanikolopoulos. A novel method for tracking andcounting pedestrians in real-time using a single camera[J]. VehicularTechnology, IEEE Transactions on.2001,50(5):1267~1278
[103]D. Kong, D. Gray, H. Tao. A viewpoint invariant approach for crowdcounting[C]. Pattern Recognition(ICPR’06).2006,3:1187~1190
[104]D. Ryan, S. Denman, S. Sridharan, C. Fookes. Scene Invariant CrowdCounting[C]. The International Conference on Digital Image Computing:Techniques and Applications.2011:237~242
[105]B.D. Lucas, T. Kanade. An iterative image registration technique with anapplication to stereo vision[C]. Proceedings of the7th international jointconference on Artificial intelligence.1981:121~130
[106]M.C. Allmen. Image sequence description using spatiotemporal flow curves:Toward motion-based recognition[D]. Wisconsin, USA: University ofWisconsin-Madison(Ph D dissertation),1991
[107]A. Kumar, A.R. Tannenbaum, G. Balas. Optical flow: a curve evolutionapproach[J]. IEEE Trons. on Image Processing.1996,5(4);598~610
[108]T. Brox, A. Bruhn, N. Papenberg, J. Weickert. High accuracy optical flowestimation based on a theory for warping[C]. Computer Vision-ECCV2004:25~36
[109]L. Kratz, K. Nishino. Tracking with local spatio-temporal motion patternsin extremely crowded scenes[C]. Computer Vision and Pattern Recognition(CVPR),2010IEEE Conference on.2010:693~700
[110]S. Zhu S, Y. Wu, D. Mumford. Minimax entropy principle and itsapplication to texture modeling[J]. Neural Computation,1997,9:1627~1660
[111]Doretto G, Chiuso A, Wu Y N, et al. Dynamic textures[J]. InternationalJournal of Computer Vision.2003,51(2):91~109
[112]A. B. Chan and N. Vasconcelos. The EM algorithm for mixtures of dynamictextures. Technical Report SVCL-TR-2005-02, http://www.svcl.ucsd.edu,UCSD,2005
[113]http://www.bbcmotiongallery.com/,2012
[114]V. Mahadevan, W. Li, V. Bhalodia V. http://www.svcl.ucsd.edu/projects/:Statistical Visual Computing Lab,2012
[115]http://www.cvg.rdg.ac.uk/PETS2009/a.html#s0: IEEE,2012
[116]http://mha.cs.umn.edu/Movies/Crowd-Activity-All.avi: ArtificalIntelligence, Robotics and Vision Laboratory of University of Minnesota,Department of Computer Science and Engineering,2012
[117]J. Kim, K. Grauman. Observe locally, infer globally: A space-time MRF fordetecting abnormal activities with incremental updates[C]. In:Proceedings of the IEEE Computer Society Conference on Computer Visionand Pattern Recognition, Florida, USA,2009:2921~2928
[118]V. Mahadevan, W. Li, V. Bhalodia. Anomaly detection in crowded scenes.In: Proceedings of the IEEE Computer Society Conference on ComputerVision and Pattern Recognition, San Francisco, USA,2010.1975–1981
[119]B.A. Boghossian, S.A. Velastin. Motion-based machine vision technique forthe management of large crowds[C]. Proc.6thIEEE Int. Conf. Electronics,Circuits and Systems.1999,2:961~964
[120]D. Helbing and H. Z. Al-Abideen. The dynamics of crowd disasters: Anempirical study[J]. Physical Review E.2007,75:046109
[121]A. Johansson, D. Helbing, H.Z. Al-Abideen, S. Al-Bosta. From crowddynamics to crowd safety: A video-based analysis[J]. Advances in ComplexSystems (ACS).{2008},11(04):497~527
[122]T. Nagatani. The physics of traffic jams[J]. Reports on Progress in Physics.2002,65. Doi:10.1088/0034-4885/65/9/203
[123]D. Helbing, A. Johansson, J. Mathiesen, M.H. Jensen, A. Hansen.Analytical approach to continuous and intermittent bottleneck flows[J].Physical review letters.2006,97(16),DOI:10.1103/PhysRevLett.97.168001
[124]W. Brendel, A. Fern, S. Todorovic. Probabilistic event logic forinterval-based event recognition[C]. Computer Vision and PatternRecognition (CVPR’11).2011:3329~3336
[125]R. Roberts, C. Potthast, F. Dellaert. Learning general optical flow subspacesfor egomotion estimation and detection of motion anomalies[C]. ComputerVision and Pattern Recognition(CVPR’09).2009:57~64
[126]M. Hu, S. Ali, M.Shah. Learning motion patterns in crowded scenes usingmotion flow field[C]. Pattern Recognition(ICPR’08).2008:1~5
[127]C. Dorin, M. Peter. Mean Shift: A Robust Approach toward Feature SpaceAnalysis[J]. IEEE Transactions on Pattern Analysis and MachineIntelligence.2002,24(5):603~619
[128]Y. Cheng. Mean shift, mode seeking, and clustering[J]. IEEE Trons. onPAMI.1995,17(8):790~799
[129]A. Martin. A metric for ARMA processes[J]. IEEE Trans. on SignalProcessing.2000,48(4):1164~1170
[130]A. Adam, E. Rivlin, Shimshoni, D. Reinitz. Robust real-time unusual eventdetection using multiple fixed-location monitors[J]. IEEE Trans. on PatternAnalysis and Machine Intelligence. March2008,30(3):555~560
[2] D. Helbing, L. Buzna, A. Johansson, et al. Self-organized Pedestrian CrowdDynamics: Experiments, Simulations, and Design Solutions[J].Transportation science.2005,39(1):1~24
[3] A.N. Marana, S.A. Velastin, L.F. Costa, R.A. Lotufo. Automatic estimationof crowd density using texture[J]. Safety Science.1998,28(3):165~175
[4] N. Dong, F. Liu, Z. Li. Crowd density estimation using sparse texturefeatures[J]. Journal of Convergence Information Technology.2010,5(6):125~137
[5] W. Li, X. Wu, K. Matsumoto, H.A. Zhao. Crowd density estimation: Animproved approach[C]. IEEE10th International Conference on SignalProcessing (ICSP’10).2010:1213~1216
[6] X. Wu, G. Liang, K.K. Lee, Y. Xu. Crowd density estimation using textureanalysis and learning[C]. IEEE International Conference on Robotics andBiomimetics(ROBIO'06).2006:214~219
[7] V. Rabaud, S. Belongie. Counting crowded moving objects[C]. ComputerVision and Pattern Recognition(CVPR’06), IEEE Computer SocietyConference on.17-22June2006, New York, NY, USA.2006,1:705~711
[8] H. Song, L. Xin, Y.Z. Chen, Y.Y. Gao. Weather identifying system basedon vehicle video image[C]. Intelligent Control and Automation (WCICA),9th World Congress on.21-25June2011, Taipei TaiWan.2011:172~175
[9] M. Roser, F. Moosmann. Classification of weather situations on single colorimages[C]. Intelligent Vehicles Symposium,2008IEEE. Eindhoven,Netherlands, June4-6,2008:798~803
[10] http://www.lyyn.com/showroom/applications/fog-and-haze/. LYYN,2012
[11] B.K.P. Horn, B.G. Schunck. Determining optical flow[J]. Artificialintelligence.1981,17(1-3):185~203
[12] A.B. Chan, Z.S.J. Liang, N. Vasconcelos. Privacy preserving crowdmonitoring: Counting people without people models or tracking[C]. IEEEConference on Computer Vision and Pattern Recognition(CVPR’08).2008:1~7
[13] S.F. Lin, J.Y. Chen, H.X. Chao. Estimation of number of people in crowdedscenes using perspective transformation[J]. Systems, Man and Cybernetics,Part A: Systems and Humans, IEEE Transactions on.2001,31(6):645~654
[14] V. Zlokolica, A. Pizurica, W. Philips. Video denoising using multiple classaveraging with multiresolution[J]. Visual Content Processing andRepresentation.2003,2849:172~179
[15] I.W. Selesnick, K.Y. Li. Video denoising using2D and3D dual-treecomplex wavelet transforms[J]. Wavelets: Applications in Signal and ImageProcessing X.2003,5207:607~618
[16] H.J. Seo, P. Milanfar. Video denoising using higher order optimalspace-time adaptation[C]. Acoustics, Speech and Signal Processing,2008.ICASSP2008. Santa Cruz, CA, March312008-April42008,1249~1252
[17] H. Ji, C. Liu, Z. Shen, Y. Xu. Robust video denoising using low rank matrixcompletion[C]. Computer Vision and Pattern Recognition (CVPR’10), IEEEConference on. June13-18,2010, San Francisco, CA, USA:1791~1798
[18] J. Chen, C.K. Tang. Spatio-temporal Markova Random Field for videodenoising[C]. In: Proc. of the IEEE Conference on Computer Vision andPattern Recognition(CVPR'07).18-23June2007, Minneapolis, Minnesota,USA.1~8
[19] F. Luisier, T. Blu, M. Unser. SURE-LET for orthonormal wavelet-domainvideo denoising[J]. IEEE Trans. Circuits and Systems for Video Technology.2010,20(6):913~919
[20] G. Varghese, Z. Wang. Video denoising based on a spatiotemporal Gaussianscale mixture model[J]. Circuits and Systems for Video Technology, IEEETransactions on.2010,20(7):1032~1040
[21] R. Eslami, X. L Wu. Video denoising using3-D Hybrid Wavelets andDirectional filter banks[C]. Image Processing2008(ICIP2008), San Diego,October12-15,2008,2340~2343
[22] C. Liu, W. Freeman. A high-quality video denoising algorithm based onreliable motion estimation[J]. Computer Visio (ECCV’10).2010:706~719
[23] V. Zlokolica, W. Philips. Motion and Detail Adaptive Denoising in Video[J].Proc. SPIE—Image Processing: Algorithms and Systems III,2004(5298):403~412
[24] H. Y. Cheong, A. Tourapis, J. Llach, J. Boyce. Adaptive spatio-temporalfiltering for video-denoising[J]. IEEE Trans. Image Processing.2004,2:965~968
[25] J. Boulanger, C. Kervrann, P. Bouthemy. Space–time adaptation forpatch-based image sequence restoration[J]. IEEE Trans. PAMI,2007(29):1096~1102
[26] S. Kindermann, S. Osher, P.W. Jones. Deblurring and denoising of imagesby nonlocal functionals[J]. Multiscale Modeling Simulation,2005(4):1091~1115
[27] A. Elmoataz, O. Lezoray, S. Bougleux. Nonlocal discrete regularization onweighted graphs: a framework for image and manifold processing[J]. IEEETrans. Image Process.2008,17(7):1047~1060
[28] M. Ghoniem, Y. Chahir, A. Elmoataz. Video denoising and simplificationvia discrete regularization on graph[C]. Advanced Concepts for IntelligentVision Systems (ACIVS’08),2008:380~389
[29] M. Ghoniem, Y. Chahir, A. Elmoataz. Video denoising via discreteregularization on graphs[C]. ICPR’08, Florida, USA, December8-11,2008:1~4
[30] M. Protter, M. Elad, H. Takeda, P. Milanfar, Generalizing the nonlocal-means to super-resolution reconstruction[J]. IEEE Trans. Image Process.2009,18(1):36~51
[31] A. Buades, B. Coll, J. M. Morel. Nonlocal image and movie denoising[J].Comput Vision,2008,76(2):123~139
[32] M. Ghoniem, Y. Chahir, A. Elmoataz. Nonlocal video denoising,simplification and inpainting using discrete regularization on graphs[J].Signal Processing.2008,90(8):2445~2455
[33] R.S. Lee, C. Craun, M. Robinson, M. Phaneuf. Evaluating the impacts ofhaze on air traffic operations[C]. The92nd American MeteorologicalSociety Annual Meeting. January22-26,2012, New Orleans, LA, USA.http://ams.confex.com/ams/92Annual/webprogram/Paper201227.html(2012)
[34] K.M. He, J. Sun, X.O Tang. Single Image Haze Removal Using DarkChannel Prior[J]. IEEE Transactions on Pattern Analysis and MachineIntelligence.2011,33(12):2043~2051
[35] S. Shwartz, E. Namer, Y.Y. Schechner. Blind Haze Separation[C]. Proc.IEEE Conf. Computer Vision and Pattern Recognition.17-22June2006,New York, NY, USA.2006,2:1984~1991
[36] X. Zhang, H. Li, Y. Qi, W.K. Leow, T.K. Ng. Rain removal in video bycombining temporal and chromatic properties[C]. Proceedings of the2006IEEE International Conference on Multimedia and Expo. Washington DC:IEEE Computer Society,2006:461~464
[37] P. Barnum, T. Kanade, S.G. Narasimhan. Spatio-temporal frequencyanalysis for removing rain and snow from videos [C]//Proceedings of theFirst International Workshop on Photometric Analysis for Computer Vision,in conjunction with International Conference of Computer Vision(ICCV2007). Rio de Janeiro: INRIA,2007:8p
[38] P. Barnum, Narasimhan S, Kanade T. Analysis of rain and snow infrequency space [J]. International Journal of Computer Vision.2009,86(2-3):256~274
[39] X. S. Kong, A. D. Zhang, Y. G. Qian. Haze and cloud cover recognition andremoval for serial Landsat images. Proc. SPIE8006,80061K (2011);doi:10.1117/12.902875(Remote Sensing Image Processing, GeographicInformation Systems, and Other Applications)
[40] F. Cogun, A.E. Cetin. Object tracking under illumination variations using2D-cepstrum characteristics of the target[C]. Multimedia Signal Processing(MMSP),2010IEEE International Workshop on.4-6Oct.2010, Saint-Malo,France.2010:521~526
[41] Freedman D, Turek M W. Illumination-invariant tracking via graph cuts [C]//Proceedings of the2005IEEE Computer Society Conference on ComputerVision and Pattern Recognition. Washington DC: IEEE Computer Society,2005(2):10~17
[42] D. Zang. Illumination invariant object tracking based on multiscalephase[C]. Image Processing (ICIP),201017th IEEE InternationalConference on.26-29Sept.2010, Shanghai, China.2010:357~360
[43] M. Haindl, P. Vacha. Illumination Invariant Texture Retrieval[C]. PatternRecognition(ICPR2006),18th International Conference on.20-24Aug.2006, Hong Kong, China.2006,3:276~279
[44]单勇，王润生.适应灰度和光照变化的运动目标跟踪方法[J].计算机辅助设计与图形学学报,2006,18(2):283~288
[45] Slater D, Healey G. The illumination-invariant recognition of3D objectsusing local color invariants [J]. IEEE Transactions on Pattern Analysis andMachine Intelligence,1996,18(2):206~210
[46] T. Hou, S. Wang, H. Qin. Vehicle matching and recognition under largevariations of pose and illumination[C]. Computer Vision and PatternRecognition Workshops,2009. CVPR Workshops2009. IEEE ComputerSociety Conference on.20-25June2009, Miami, FL, USA.2009:24~29
[47] P. Felzenszwalb, R. Girshick, D. McAllester, D. Ramanan. Object detectionwith discriminatively trained part based models[J]. IEEE Trans. PAMI.2010,32(9):1627~1645
[48] R. Mehran, A. Oyama, M. Shah. Abnormal crowd behavior detection usingsocial force model[C]. Computer Vision and Pattern Recognition(CVPR’09)IEEE Conference on.20-25June2009, Miami, FL.2009:935~942
[49] M. Rodriguez, I. Laptev, J. Sivic, J.Y. Audibert. Density-aware persondetection and tracking in crowds[C]. Computer Vision (ICCV’11), IEEEInternational Conference on.6-13Nov.2011, Barcelona, Spain.2011:2423~2430
[50] O. Tuzel, F. Porikli, P. Meer. Pedestrian Detection via Classification onRiemannian Manifolds[J]. IEEE Trans. PAMI.2008,30(10):1713~1727
[51] S. Wang, Z.J. Miao. Anomaly detection in crowd scene using historicalinformation[C]. Intelligent Signal Processing and Communication Systems(ISPACS), International Symposium on.2010:1~4
[52] S. Stalder, H. Grabner, L.V. Gool. Exploring context to learn scene specificobject detectors[C]. Proceedings11th IEEE International Workshop onPETS. Miami, June25,2009:63~70
[53] L. Kratz, K. Nishino. Anomaly detection in extremely crowded scenes usingspatio-temporal motion pattern models[C]. Computer Vision and PatternRecognition(CVPR’09). IEEE Conference on,2009:1446~1453
[54] K.M. Zeitz, H.M. Tan, M. Grief, P.C. Couns, C.J. Zeitz. Crowd behavior atmass gatherings: a literature review[J]. Prehospital and disaster medicine.2010,24(1):32~38
[55] M. Mancas. Attention-Based Dense Cropwds Analysis[C]. Image Analysisfor Multimedia Interactive Services (WIAMIS), the11th InternationalWorkshop on.2010:1~4
[56] D. Ryan, S. Denman, C. Fookes, S. Sridharan. Crowd counting usingmultiple local features[J]. Digital Image Computing: Techniques andApplications.2009,81~88
[57] S. Denman, V. Chandran, and S. Sridharan. An adaptive optical flowtechnique for person tracking systems[J]. Elsivier Pattern RecognitionLetters.2007,28(10):1232~1239
[58] W. Ge R. Collins, R. Ruback. Vision-based analysis of small groups inpedestrian crowds[J]. IEEE Trons. PAMI.2011,99:1~14
[59] M.C. Chang, N. Krahnstoever, S. Lim, Y.T. Group Level ActivityRecognition in Crowded Environments across Multiple Cameras[C].Proceedings of the7th IEEE International Conference on Advanced Videoand Signal Based Surveillance (AVSS’10),29August-1September2010,Boston, Massachusetts.2010:56~63
[60] S. Ali, M. Shah. A Lagrangian particle dynamics approach for crowd flowsegmentation and stability analysis[C]. Computer Vision and PatternRecognition(CVPR'07). IEEE Conference on.2007:1~6
[61] A.B. Chan, N. Vasconcelos. Modeling, clustering, and segmenting videowith mixtures of dynamic textures[J]. IEEE Trons. PAMI.2008,30(5):909~926
[62] E.L. Andrade, R.B. Fisher. Simulation of crowd problems for computervision[C]. First International Workshop on Crowd Simulation.2005,3:71~80
[63] N. Pelechano, J.M. Allbeck, N.I. Badler. Controlling individual agents inhigh-density crowd simulation[C]. Proceedings of the2007ACMSIGGRAPH/Eurographics symposium on Computer animation.2007:99~108
[64] C. Peters, C. Ennis. Modeling groups of plausible virtual pedestrians[J].Computer Graphics and Applications.2009,29(4):54~63
[65] R A. Skov. Impacts of haze on high-capacity airport operations. The14thConference on Aviation, Range, and Erospace Meteorology. NOAA/NWS,Hampton, GA,2010
[66] Medina, Juan C, Chitturi, Madhav, Benekohal, Rahim F. Effects of fog,snow, and rain on video detection systems at intersections. TransportationLetters, International Journal of Transportation Research.2010,2(1):1~12
[67] X.L. Yu, C. B Xiao, M. Deng, P. Li. A Classification Algorithm toDistinguish Image as Haze or Non-haze[C]. In Proceedings of the SixthInternational Conference on Image and Graphics (ICIG’11). Hefei, China.August12-15,2011:286~289
[68] A. Elgammal, D. Harwood, L. Davis. Non-parametric model for backgroundsubtraction[C]. Computer Vision(ECCV2000, Springer). Dublin, Ireland,June26-July1,2000:751~767
[69] M.R. Blanton, E. Kazin, D. Muna, B.A. Weaver, A. Price-Whelan.Improved background subtraction for the Sloan Digital Sky Surveyimages[J]. The Astronomical Journal.2011,142:31pages, IOP Publishing.doi:10.1088/0004-6256/142/1/31
[70] Y. Benezeth, P.M. Jodoin, B. Emile, H. Laurent, C. Rosenberger.Comparative study of background subtraction algorithms[J]. Journal ofElectronic Imaging.2010,19:033003
[71] A. Lipton, H. Fujiyoshi, R. Patil. Moving target classification and trackingfrom real-time video[C]. In: Proc IEEE Workshop on Applications ofComputer Vision, Princeton, NJ,1998:8~14
[72] á. Bayona, J. SanMiguel, J. Martinez. Stationary foreground detectionusing background subtraction and temporal difference in videosurveillance[C]. Image Processing (ICIP’10), the17th IEEE InternationalConference on. September26-292010, Hong Kong, China,2010:4657~4660
[73] J.L. Barron, D.J. Fleet, S.S. Beauchemin. Performance of optical flowtechniques[J]. International journal of computer vision.1994,12(1):43~77
[74] S. Baker, D. Scharstein, J.P. Lewis, S. Roth,M.J. Black, R. Szeliski. Adatabase and evaluation methodology for optical flow[J]. InternationalJournal of Computer Vision.2011,92(1):1~31
[75] M.B. Holte, T.B. Moeslund, P. Fihl. View-invariant gesture recognitionusing3D optical flow and harmonic motion context[J]. Computer Visionand Image Understanding.2010,114:1353~1361
[76] K.P. Murphy. Dynamic Bayesian networks[J]. Probabilistic GraphicalModels, in M. Jordan.2002. http://www.cs.ubc.ca/~murphyk/Papers/dbn-chapter.pdf (2012)
[77] Z. Zeng, Q. Ji. Knowledge based activity recognition with dynamicBayesian network[J]. Computer Vision.2010:532~546
[78] H.I. Suk, B.K. Sin, S.W. Lee. Hand gesture recognition based on dynamicBayesian network framework[J]. Pattern Recognition,2010,43(9):3059~3072
[79] J. Cortes. Distributed Kriged Kalman Filter for SpatialEstimation[J]. Automatic Control, IEEE Transactions on.2009,12:2816~2827
[80] M. Mehta, C. Goyal, M.C. Srivastava, R.C Jain. Real time object detectionand tracking: Histogram matching and Kalman filter approach[C]. Computerand Automation Engineering (ICCAE’10), The2nd InternationalConference on.26-28Feb.2010, Singapore.2010,5:796~801
[81] H. Medeiros, J. Park, A. Kak. Distributed Object Tracking Using aCluster-Based Kalman Filter in Wireless Camera Networks[J]. IEEE Journalof Selected Topics in Signal Processing.2008,2(4):448~463
[82] Y.M. Lui, J.R. Beveridge, L.D. Whitley. A novel appearance model andadaptive condensation algorithm for human face tracking[C]. Biometrics:Theory, Applications and Systems(BTAS’08), the2nd IEEE InternationalConference on. Sept.29-Oct.1,2008, Arlington, VA:1~7
[83] L. Dan, O. Jun. Hand-gesture extraction and recognition from the videosequence acquired by a dynamic camera using condensation algorithm[C].Proc. SPIE7252,72520S (2009);19January2009, San Jose, CA, USA.http://dx.doi.org/10.1117/12.806108(2012)
[84] N. Hauti`ere, R. Labayrade, D. Aubert. Real-time disparity contrastcombination for onboard estimation of the visibility distance[J]. IEEETransactions on Intelligent Transportation Systems.2006(7):201~212
[85] E. Dumont, V. Cavallo. Extended photometric model of fog effects on roadvision[J]. Transp. Transportation Research Record: Journal of theTransportation Research Board.2004(1862):77~81
[86] H. Koschmieder. Theorie der Horizontalen Sichtweite. Beitr. Phys. FreienAtm.1924,12:171~181
[87] Narasimhan S G, Nayar S K. Vision and the Atmosphere[J]. InternationalJournal of Computer Vision.2002,48(3):233~254
[88] K. Miller. Least squares method for ILL-posed problems with a prescribedbound[J]. International Journal of Math Analysis.1990,(1):52~74
[89] M. Egmont-Petersen, D. de Ridder, H. Handels. Image processing withneural networks--a review[J]. J. of Pattern Recognition.2002,(35):2279~2301
[90] S.Z. Li. Markov Random Field Modeling in Image Analysis(ThirdEdition)[M]. Springer-Verlag New York Inc.2009.371p
[91] R. Rajagopalan. Synthesizing processed video by filtering temporalrelationships. IEEE Trans. on Image Processing.2002,11(1):26~36
[92] M. Mahmoudi, G. Sapiro. Fast image and video denoising via nonlocalmeans of similar neighborhoods[J]. Signal Processing Letters, IEEE.2005,12(12):839~842
[93] H. Faraji, M. J. MacLean. CCD noise removal in digital image[J]. IEEETransactions on Image Processing.2006,15(9):2676~2685
[94] K. Garg, S.K. Nayar. Vision and rain [J]. International Journal of ComputerVision,2007,75(1):3~27
[95] K. Garg, S. K. Nayar. Detection and removal of rain from videos[C]//Proceedings of the2004IEEE Computer Society Conference on ComputerVision and Pattern Recognition. Washington DC: IEEE Computer Society,2004(1):528~535
[96] K. Garg, S. K. Nayar. Photorealistic rendering of rain streaks[J]. ACMTransactions on Graphics,2006,25(3):996~1002
[97] Brewer N, Liu N. Using the Shape Characteristics of Rain to Identify andRemove Rain from Video[J]. Structural, Syntactic, and Statistical PatternRecognition.2008,5342:451~458
[98] G. B. Foote, P. S. Du Toit. Terminal Velocity of Raindrops Aloft[J]. Journalof Applied Meteorology.1969,8(2):249~253
[99] B.K.P. Horn, B.G. Schunck. Determining optical flow[J]. Artificialintelligence.1981,17(1-3):185~203
[100] A.B. Chan, N. Vasconcelos. Mixtures of dynamic textures[C].Computer Vision(ICCV’05).2005,1:641~647
[101]V. Lempitsky, A. Zisserman. Learning To Count Objects in Images[J]. InNIPS2010:1~9
[102]O. Masoud, N.P. Papanikolopoulos. A novel method for tracking andcounting pedestrians in real-time using a single camera[J]. VehicularTechnology, IEEE Transactions on.2001,50(5):1267~1278
[103]D. Kong, D. Gray, H. Tao. A viewpoint invariant approach for crowdcounting[C]. Pattern Recognition(ICPR’06).2006,3:1187~1190
[104]D. Ryan, S. Denman, S. Sridharan, C. Fookes. Scene Invariant CrowdCounting[C]. The International Conference on Digital Image Computing:Techniques and Applications.2011:237~242
[105]B.D. Lucas, T. Kanade. An iterative image registration technique with anapplication to stereo vision[C]. Proceedings of the7th international jointconference on Artificial intelligence.1981:121~130
[106]M.C. Allmen. Image sequence description using spatiotemporal flow curves:Toward motion-based recognition[D]. Wisconsin, USA: University ofWisconsin-Madison(Ph D dissertation),1991
[107]A. Kumar, A.R. Tannenbaum, G. Balas. Optical flow: a curve evolutionapproach[J]. IEEE Trons. on Image Processing.1996,5(4);598~610
[108]T. Brox, A. Bruhn, N. Papenberg, J. Weickert. High accuracy optical flowestimation based on a theory for warping[C]. Computer Vision-ECCV2004:25~36
[109]L. Kratz, K. Nishino. Tracking with local spatio-temporal motion patternsin extremely crowded scenes[C]. Computer Vision and Pattern Recognition(CVPR),2010IEEE Conference on.2010:693~700
[110]S. Zhu S, Y. Wu, D. Mumford. Minimax entropy principle and itsapplication to texture modeling[J]. Neural Computation,1997,9:1627~1660
[111]Doretto G, Chiuso A, Wu Y N, et al. Dynamic textures[J]. InternationalJournal of Computer Vision.2003,51(2):91~109
[112]A. B. Chan and N. Vasconcelos. The EM algorithm for mixtures of dynamictextures. Technical Report SVCL-TR-2005-02, http://www.svcl.ucsd.edu,UCSD,2005
[113]http://www.bbcmotiongallery.com/,2012
[114]V. Mahadevan, W. Li, V. Bhalodia V. http://www.svcl.ucsd.edu/projects/:Statistical Visual Computing Lab,2012
[115]http://www.cvg.rdg.ac.uk/PETS2009/a.html#s0: IEEE,2012
[116]http://mha.cs.umn.edu/Movies/Crowd-Activity-All.avi: ArtificalIntelligence, Robotics and Vision Laboratory of University of Minnesota,Department of Computer Science and Engineering,2012
[117]J. Kim, K. Grauman. Observe locally, infer globally: A space-time MRF fordetecting abnormal activities with incremental updates[C]. In:Proceedings of the IEEE Computer Society Conference on Computer Visionand Pattern Recognition, Florida, USA,2009:2921~2928
[118]V. Mahadevan, W. Li, V. Bhalodia. Anomaly detection in crowded scenes.In: Proceedings of the IEEE Computer Society Conference on ComputerVision and Pattern Recognition, San Francisco, USA,2010.1975–1981
[119]B.A. Boghossian, S.A. Velastin. Motion-based machine vision technique forthe management of large crowds[C]. Proc.6thIEEE Int. Conf. Electronics,Circuits and Systems.1999,2:961~964
[120]D. Helbing and H. Z. Al-Abideen. The dynamics of crowd disasters: Anempirical study[J]. Physical Review E.2007,75:046109
[121]A. Johansson, D. Helbing, H.Z. Al-Abideen, S. Al-Bosta. From crowddynamics to crowd safety: A video-based analysis[J]. Advances in ComplexSystems (ACS).{2008},11(04):497~527
[122]T. Nagatani. The physics of traffic jams[J]. Reports on Progress in Physics.2002,65. Doi:10.1088/0034-4885/65/9/203
[123]D. Helbing, A. Johansson, J. Mathiesen, M.H. Jensen, A. Hansen.Analytical approach to continuous and intermittent bottleneck flows[J].Physical review letters.2006,97(16),DOI:10.1103/PhysRevLett.97.168001
[124]W. Brendel, A. Fern, S. Todorovic. Probabilistic event logic forinterval-based event recognition[C]. Computer Vision and PatternRecognition (CVPR’11).2011:3329~3336
[125]R. Roberts, C. Potthast, F. Dellaert. Learning general optical flow subspacesfor egomotion estimation and detection of motion anomalies[C]. ComputerVision and Pattern Recognition(CVPR’09).2009:57~64
[126]M. Hu, S. Ali, M.Shah. Learning motion patterns in crowded scenes usingmotion flow field[C]. Pattern Recognition(ICPR’08).2008:1~5
[127]C. Dorin, M. Peter. Mean Shift: A Robust Approach toward Feature SpaceAnalysis[J]. IEEE Transactions on Pattern Analysis and MachineIntelligence.2002,24(5):603~619
[128]Y. Cheng. Mean shift, mode seeking, and clustering[J]. IEEE Trons. onPAMI.1995,17(8):790~799
[129]A. Martin. A metric for ARMA processes[J]. IEEE Trans. on SignalProcessing.2000,48(4):1164~1170
[130]A. Adam, E. Rivlin, Shimshoni, D. Reinitz. Robust real-time unusual eventdetection using multiple fixed-location monitors[J]. IEEE Trans. on PatternAnalysis and Machine Intelligence. March2008,30(3):555~560
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |