面向图像处理的概率图模型应用研究
|
摘要
在图像信息智能分析中，不确定性问题的研究成为一项重要而又具有挑战性的工作，概率图模型为解决这一问题提供了一种重要途径。本文对基于概率图模型的图像局部特征描述、非匀质图像分割和视觉跟踪的问题进行了深入研究。
图像局部特征描述的核心问题是不变性(鲁棒性)和可区分性。然而特征描述子的可区分性的强弱和其不变性是矛盾的，也就是说，一个具有诸多不变性的特征描述子，它区分图像局部内容的能力就稍弱；而一个非常容易区分不同图像局部内容的特征描述子，它的鲁棒性往往比较低。因此，在基于局部特征的图像处理中，研究不仅具有较强不变性、还具有较好区分性的特征描述子有着重要意义。
非匀质图像是包含有多个灰度级目标的图像。传统方法中的边界模型虽然能够分割非匀质图像，但对噪声、弱边缘或不连续边较敏感，而区域模型中分片匀质的统计假设，使其无法分割灰度非匀质图像。因此，研究基于概率图模型融合局部特征信息的区域模型，对非匀质图像进行快速分割的问题，有着较重要的学术价值和应用前景。
视觉跟踪过程中，目标表观的动态变化容易造成目标漂移或丢失问题。传统的模型更新算法大多采用固定的先验模型克服该问题。一方面，普适的固定先验容易使被跟踪目标漂移到相似的对象；另一方面，受限的固定先验因无法处理突变的目标表观而导致跟踪失败。因此，在跟踪过程中，利用概率图模型解决不确定性信息的能力，构建在线自适应先验表观模型是具有良好应用前景的研究方向。
论文取得的主要成果与创新工作概括如下：
①在形状变化、被遮挡、噪声情况下，存在难以完整分割目标形状的问题，为此，提出一种采用隐含形状约束马尔可夫随机场模型的形状分割方法。引入目标的先验形状知识，用水平集符号距离函数隐含表示目标的先验形状模型；以先验形状模型作为约束构造出MRF能量函数；采用graph cut法求解能量函数极小值，利用形状对准和最大流法演化初始轮廓，快速准确地分割出目标的形状。实验表明新方法能有效快速分割带遮挡、噪声以及发生形状变化的目标，增强了形状分割的鲁棒性。
②针对灰度非匀质图像分割困难及效率低下的问题，提出了一种基于马尔可夫随机场的局部区域活动轮廓模型快速分割方法。该方法结合核函数和割测度定义一个新的能量函数。一方面，在中心点被核函数掩模的局部区域内，用邻近点的加权均值拟合数据项，能有效处理图像的非匀质分布。另一方面，用割测度逼近的曲线长度作为全局正则性，利于轮廓快速定位于物体边界。在轮廓演化过程中，使用基于栅格图的最大流算法，避免了传统模型计算代价高昂的水平集函数。合成图像和真实图像的实验结果表明，提出的方法能有效快速地分割灰度非匀质图像中的弱边缘物体及复杂的多灰阶结构物体；对初始轮廓线位置和噪声具有较好的鲁棒性。
③针对目标跟踪过程中，可变目标表观的特征数据会发生“分布漂移”的问题，基于时间Dirichlet过程混合模型，提出了一种学习多模表观模型的方法并用于目标跟踪。以时间Dirichlet过程为先验分布，把先前估计的目标样本划分为不同的聚集，使得每个聚集表示一类表观，每个表观类被建模为判别式分类器；基于贝叶斯后验推断，权衡先前表观模型的分类误差和拆分聚集的代价，从数据中自主学习表观模型；基于Noisy-OR模型，以贪心（Greedy）策略协同各表观分类器判别出目标。仿真结果表明该方法能较好的跟踪可变目标表观，改善了目标跟踪性能。
④为有效解决视觉跟踪过程中的目标“漂移”问题，基于分层Dirichlet过程演化聚类模型，提出一种采用自适应先验表观模型的目标跟踪方法。该方法在一致的架构内融合HDP-EVO演化聚类模型和在线Boosting学习。以分层Dirichlet过程为先验分布，对总体表观示例进行聚类分析，获得随时间自适应演化的表观类先验知识；另一方面，利用共享的表观类混合比例的权重平滑约束各时刻的表观模型；改进Gibbs抽样过程，使之能融入目标示例的分类误差，并交替迭代地从数据中自主学习聚类和表观分类器；根据表观模型中各表观类的权重系数组合它们的分类评分去定位目标位置。仿真实验表明新方法学习的表观模型能较鲁棒的自适应于目标的表观变化，提高了目标跟踪精度。
Information uncertainty in image processing has become a new hot-spot as achallenging research. Probabilistic graphical model provides an important means forresolving the uncertainty of intelligent information field. This paper made an intensivestudy on local features of an image, segmentation for the in-homogenous images andvisual tracking based upon probabilistic graphical model.
Robustness and discrimination are the core problems of different kinds of localfeature descriptors, but robustness and discrimination are contradictory. Namely, for afeature descriptor with robustness, its power of discriminating the content of an image isweak. Correspondingly, the one with discrimination has worse robustness. Hence, tobalance the trade-off between robustness and discrimination, the research based on localfeature descriptors has very important value in mage processing.
Image with intensity in-homogeneity consists of multiple gray levels. Althoughclassical edge-based models can segment the image with intensity in-homogeneity, theyare sensitive for segmenting the objects with weak or discontinuous boundary and noise;region-based models have been successfully used in binary phase segmentation with theassumption that each image region is statistically homogeneous. However, region-basedmodels do not work well for the image with intensity in-homogeneity. Hence, theresearch of region-based models about integrating local feature information based onprobabilistic graphical model has very important academic and application role for thein-homogeneous image.
Appearance variations of the target object will cause the drifting or missingproblem of the target object during tracking. To cope with the problem, most existingonline methods utilize fixed prior model to update appearance model. These methodsare either too generic to drift to the similar objects or too restrictive to fail in dramaticchanges. Therefore, a promising research direction is to construct an adaptive priormodel which can adapt to changes incrementally during tracking.
The main contributions of this paper are summarized as follows:
First of all, to deal with shape segmentation of the object in complicated scenes, anovel method based on MRF model with implicit shape prior was proposed. First, theobject was segmented using the initial contour. Then, an implicit shape model with levelset signed distance function was built to improve the accuracy of object segmentation, and a MRF energy function with the constraint of the existing shape prior wasconstructed. The optimal value of energy function could be found by graph cut method.The precise segmentation of object was obtained by applying the shape alignment andmax-flow method to evolve the initial contour. Experimental results show that theproposed method can better cope with the clutter and noise, as well as partial occlusionsand affine transformation of the shape. Thus the robust stability of segmentation resultis improved.
Secondly, to solve effectively the difficult and ineffective segmentation for thein-homogenous images, a novel fast method based on the local region active contourmodel was proposed. A new energy function was defined by combining kernel functionand cut metric. Utilization of the kernel function was favor of computing thein-homogenous distribution of local regions effectively. On the other hand, betterapproximation of the curve length by cut metric could help the contours to evolve intothe object boundary quickly. In addition, in the evolving process of contours, amax-flow method was adopted, which avoided an expensive computational level setmethod. Experimental results using synthetic and real images show that the proposedmethod can effectively segment objects with the weak boundary in in-homogenousimages, as well as the complex structure objects with multi-gray levels. At the sametime, the method is robust to noise and the initial contours.
Thirdly, to cope with appearance variations of the target object during visualtracking, a nonparametric Bayesian multi-modal appearance model for learning overtime was proposed. First, by taking the temporal Dirichlet process as prior distribution,the proposed model separated target samples previously estimated into several clusters.Each cluster represented a certain type of the target appearance, which was modeled asdiscriminative classifiers. Then, to balance the trade-off between the classification errorof appearance model and the cost for splitting the clusters, the multi-modal appearancemodel was automatically learned by the use of Bayesian posterior inference. Finally,based on the Noisy-OR model, a greedy algorithm was used to discriminate the targetobject by combining the outputs of appearance classifiers. The simulation results showthat the proposed method can robustly track an object under rapid appearance changesand improve the tracking performance.
Fourthly, to effectively solve the drifting problem of a varying target object duringtarget tracking, an adaptive prior appearance model was presented. First, the methodcombined hierarchical dirichlet process evolutionary clustering model and online boosting learning into a coherent framework. By taking the hierarchical dirichletprocess as prior distribution, the prior appearance knowledge could adapt to changeover time. On the other hand, appearance model of each moment was smoothlyconstrained by the mixture proportion of a type of appearance cluster. Then, to balancethe classification error of appearance model and the cost for splitting the clusters, themulti-modal appearance model was automatically learned by the use of Bayesianposterior inference. Finally, based on the weighting factor of appearance clusters, thetarget object was discriminated by combining the outputs of appearance classifiers. Thesimulation results show that the learned appearance model can adapt to the appearancevariations of the target object and achieve better tracking results with high accuracy.
图像局部特征描述的核心问题是不变性(鲁棒性)和可区分性。然而特征描述子的可区分性的强弱和其不变性是矛盾的，也就是说，一个具有诸多不变性的特征描述子，它区分图像局部内容的能力就稍弱；而一个非常容易区分不同图像局部内容的特征描述子，它的鲁棒性往往比较低。因此，在基于局部特征的图像处理中，研究不仅具有较强不变性、还具有较好区分性的特征描述子有着重要意义。
非匀质图像是包含有多个灰度级目标的图像。传统方法中的边界模型虽然能够分割非匀质图像，但对噪声、弱边缘或不连续边较敏感，而区域模型中分片匀质的统计假设，使其无法分割灰度非匀质图像。因此，研究基于概率图模型融合局部特征信息的区域模型，对非匀质图像进行快速分割的问题，有着较重要的学术价值和应用前景。
视觉跟踪过程中，目标表观的动态变化容易造成目标漂移或丢失问题。传统的模型更新算法大多采用固定的先验模型克服该问题。一方面，普适的固定先验容易使被跟踪目标漂移到相似的对象；另一方面，受限的固定先验因无法处理突变的目标表观而导致跟踪失败。因此，在跟踪过程中，利用概率图模型解决不确定性信息的能力，构建在线自适应先验表观模型是具有良好应用前景的研究方向。
论文取得的主要成果与创新工作概括如下：
①在形状变化、被遮挡、噪声情况下，存在难以完整分割目标形状的问题，为此，提出一种采用隐含形状约束马尔可夫随机场模型的形状分割方法。引入目标的先验形状知识，用水平集符号距离函数隐含表示目标的先验形状模型；以先验形状模型作为约束构造出MRF能量函数；采用graph cut法求解能量函数极小值，利用形状对准和最大流法演化初始轮廓，快速准确地分割出目标的形状。实验表明新方法能有效快速分割带遮挡、噪声以及发生形状变化的目标，增强了形状分割的鲁棒性。
②针对灰度非匀质图像分割困难及效率低下的问题，提出了一种基于马尔可夫随机场的局部区域活动轮廓模型快速分割方法。该方法结合核函数和割测度定义一个新的能量函数。一方面，在中心点被核函数掩模的局部区域内，用邻近点的加权均值拟合数据项，能有效处理图像的非匀质分布。另一方面，用割测度逼近的曲线长度作为全局正则性，利于轮廓快速定位于物体边界。在轮廓演化过程中，使用基于栅格图的最大流算法，避免了传统模型计算代价高昂的水平集函数。合成图像和真实图像的实验结果表明，提出的方法能有效快速地分割灰度非匀质图像中的弱边缘物体及复杂的多灰阶结构物体；对初始轮廓线位置和噪声具有较好的鲁棒性。
③针对目标跟踪过程中，可变目标表观的特征数据会发生“分布漂移”的问题，基于时间Dirichlet过程混合模型，提出了一种学习多模表观模型的方法并用于目标跟踪。以时间Dirichlet过程为先验分布，把先前估计的目标样本划分为不同的聚集，使得每个聚集表示一类表观，每个表观类被建模为判别式分类器；基于贝叶斯后验推断，权衡先前表观模型的分类误差和拆分聚集的代价，从数据中自主学习表观模型；基于Noisy-OR模型，以贪心（Greedy）策略协同各表观分类器判别出目标。仿真结果表明该方法能较好的跟踪可变目标表观，改善了目标跟踪性能。
④为有效解决视觉跟踪过程中的目标“漂移”问题，基于分层Dirichlet过程演化聚类模型，提出一种采用自适应先验表观模型的目标跟踪方法。该方法在一致的架构内融合HDP-EVO演化聚类模型和在线Boosting学习。以分层Dirichlet过程为先验分布，对总体表观示例进行聚类分析，获得随时间自适应演化的表观类先验知识；另一方面，利用共享的表观类混合比例的权重平滑约束各时刻的表观模型；改进Gibbs抽样过程，使之能融入目标示例的分类误差，并交替迭代地从数据中自主学习聚类和表观分类器；根据表观模型中各表观类的权重系数组合它们的分类评分去定位目标位置。仿真实验表明新方法学习的表观模型能较鲁棒的自适应于目标的表观变化，提高了目标跟踪精度。
Information uncertainty in image processing has become a new hot-spot as achallenging research. Probabilistic graphical model provides an important means forresolving the uncertainty of intelligent information field. This paper made an intensivestudy on local features of an image, segmentation for the in-homogenous images andvisual tracking based upon probabilistic graphical model.
Robustness and discrimination are the core problems of different kinds of localfeature descriptors, but robustness and discrimination are contradictory. Namely, for afeature descriptor with robustness, its power of discriminating the content of an image isweak. Correspondingly, the one with discrimination has worse robustness. Hence, tobalance the trade-off between robustness and discrimination, the research based on localfeature descriptors has very important value in mage processing.
Image with intensity in-homogeneity consists of multiple gray levels. Althoughclassical edge-based models can segment the image with intensity in-homogeneity, theyare sensitive for segmenting the objects with weak or discontinuous boundary and noise;region-based models have been successfully used in binary phase segmentation with theassumption that each image region is statistically homogeneous. However, region-basedmodels do not work well for the image with intensity in-homogeneity. Hence, theresearch of region-based models about integrating local feature information based onprobabilistic graphical model has very important academic and application role for thein-homogeneous image.
Appearance variations of the target object will cause the drifting or missingproblem of the target object during tracking. To cope with the problem, most existingonline methods utilize fixed prior model to update appearance model. These methodsare either too generic to drift to the similar objects or too restrictive to fail in dramaticchanges. Therefore, a promising research direction is to construct an adaptive priormodel which can adapt to changes incrementally during tracking.
The main contributions of this paper are summarized as follows:
First of all, to deal with shape segmentation of the object in complicated scenes, anovel method based on MRF model with implicit shape prior was proposed. First, theobject was segmented using the initial contour. Then, an implicit shape model with levelset signed distance function was built to improve the accuracy of object segmentation, and a MRF energy function with the constraint of the existing shape prior wasconstructed. The optimal value of energy function could be found by graph cut method.The precise segmentation of object was obtained by applying the shape alignment andmax-flow method to evolve the initial contour. Experimental results show that theproposed method can better cope with the clutter and noise, as well as partial occlusionsand affine transformation of the shape. Thus the robust stability of segmentation resultis improved.
Secondly, to solve effectively the difficult and ineffective segmentation for thein-homogenous images, a novel fast method based on the local region active contourmodel was proposed. A new energy function was defined by combining kernel functionand cut metric. Utilization of the kernel function was favor of computing thein-homogenous distribution of local regions effectively. On the other hand, betterapproximation of the curve length by cut metric could help the contours to evolve intothe object boundary quickly. In addition, in the evolving process of contours, amax-flow method was adopted, which avoided an expensive computational level setmethod. Experimental results using synthetic and real images show that the proposedmethod can effectively segment objects with the weak boundary in in-homogenousimages, as well as the complex structure objects with multi-gray levels. At the sametime, the method is robust to noise and the initial contours.
Thirdly, to cope with appearance variations of the target object during visualtracking, a nonparametric Bayesian multi-modal appearance model for learning overtime was proposed. First, by taking the temporal Dirichlet process as prior distribution,the proposed model separated target samples previously estimated into several clusters.Each cluster represented a certain type of the target appearance, which was modeled asdiscriminative classifiers. Then, to balance the trade-off between the classification errorof appearance model and the cost for splitting the clusters, the multi-modal appearancemodel was automatically learned by the use of Bayesian posterior inference. Finally,based on the Noisy-OR model, a greedy algorithm was used to discriminate the targetobject by combining the outputs of appearance classifiers. The simulation results showthat the proposed method can robustly track an object under rapid appearance changesand improve the tracking performance.
Fourthly, to effectively solve the drifting problem of a varying target object duringtarget tracking, an adaptive prior appearance model was presented. First, the methodcombined hierarchical dirichlet process evolutionary clustering model and online boosting learning into a coherent framework. By taking the hierarchical dirichletprocess as prior distribution, the prior appearance knowledge could adapt to changeover time. On the other hand, appearance model of each moment was smoothlyconstrained by the mixture proportion of a type of appearance cluster. Then, to balancethe classification error of appearance model and the cost for splitting the clusters, themulti-modal appearance model was automatically learned by the use of Bayesianposterior inference. Finally, based on the weighting factor of appearance clusters, thetarget object was discriminated by combining the outputs of appearance classifiers. Thesimulation results show that the learned appearance model can adapt to the appearancevariations of the target object and achieve better tracking results with high accuracy.
引文
[1] Chan F Tong, Shen Jian Hong. Image processing and analysis [M].北京:科学出版社,2009.
[2] LANKTON M S. localized statistical model in computer vision [D]. USA. Georgia Instituteof Technology,2009.
[3] Koller D, Friedman N. Probabilistic Graphical Models Principles and Technique [M].Cambridge, MA: MIT Press,2009.
[4]高琳,唐鹏,盛鹏.基于概率图模型目标建模的视觉跟踪算法[J].光电子·激光,2010,21(1):124–129.
[5] LI J, ALLINSON N M. A comprehensive review of current local features for computer vision[J]. Neurocomputing,2008,71:1771–1787.
[6] SCHELTEN K, ROTH S. Connecting non-quadratic variational models and MRFs [C]. In:Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, ColoradoSprings, USA: IEEE,2011.2641–2641.
[7] ZHANG L, JI Q. Image segmentation with a unified graphical model [J]. IEEE Transactionson Pattern Analysis and Machine Intelligence,2010,32(8):1406–1425.
[8] WU T, BAE M H, ZHANG M, PAN R, BADEA A. A prior feature SVM-MRF basedmethod for mouse brain segmentation [J]. Neuroimage,2012,59(3):2298–2306.
[9]卢易苏,陈武凡.非参数图像分割算法的研究与改进[J].计算机工程,2012,38(7):179–181.
[10] GRABNER H, LEISTNER C, BISCHOF H. Semi-supervised online boosting for robust tracking [C]. In: Proceedings of the10th European Conference on Computer Vision. Marseille, France:2008.234–247.
[11] BABENKO B, YANG M H, BELONGIE S. Robust object tracking with online multiple instance learning[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2011,33(8):1619–1632.
[12] SANTNER J, LEISTNER C, SAFFARI A, POCK T, BISCHOF H. PROST parallel robustonline simple tracking [C]. In: Proceedings of the IEEE Conference on Computer Vision andPattern Recognition. San Francisco, USA: IEEE,2010.723–730.
[13] PAN S J, YANG Q. A survey on transfer learning [J]. IEEE Transactions on Knowledge and Data Engineering,2010,22(10):1345–1359.
[14] WHITTAKER J. Graphical models in applied multivariate statistics [M]. New York: JohnWiley and Sons lnc,1990.
[15] JORDAN M I, GHAHRAMANI Z, JAAKKOLA T, SAUL L K. An introduction to variationalmethods for graphical models [J]. Machine Learning,1999,37(2):183–233.
[16] GILKS W R, RICHARDSON S, SPIEGELHALTER D J. Markov chain Monte Carlo in practice [M]. London: Chapman and Hall,1996.
[17] PEARL J. Probabilistic reasoning in intelligent systems: networks of plausible inference [M].San Mateo, California, USA: Morgan Kaufmann Publishers Inc,1988.
[18]朱允刚.贝叶斯网学习中若干问题研究及其在信息融合中的应用[D].长春:吉林大学,2012.
[19]王相海,陈明莹,方玲玲.概率图模型及其图像与视频应用研究[J].中国图象图形学报,2009,14(9):1712–1720.
[20]盛骤,谢式千,潘承毅.概率论与数理统计[M].浙江:高等教育出版社,2003.
[21] PEARL J. Bayes and Markov networks: a comparison of two graphical representations of probabilistic knowledge [R].860024(R46), Los Angeles, California, USA: University of California at Los Angeles,1986.
[22]刘李漫.基于变分模型和图割优化的图像与视频目标分割方法研究[D].武汉:华中科技大学,2012.
[23] COMANICIU D, MEER P. Mean shift: a robust approach toward feature space analysis [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(5):603–619.
[24] KASS M, WITKIN A, TERZOPOULOS D. Snakes: active contour models [J]. International Journal of Computer Visio,1988,1:321–331.
[25] SHI J, MALIK J. Normalized cuts and image segmentation [J]. IEEE Transactions on PatternAnalysis and Machine Intelligence,2000,22(8):888–905.
[26] CHAN T, VESE L. Active contours without edges [J]. IEEE Transactions on Image Processing,2001,10(2):266–277.
[27] LI STAN Z. Markov random field modeling in image analysis [M]. Third edition. Springer,2009.
[28] Lafferty J, McCallum A, Pereira F. Conditional random fields: probabilistic models forsegmention and labeling sequence data[C]. In: Proceedings of Machine Learning, Freiburg,GERMANY,2001.282–289.
[29] NIKOU C, LIKAS A C, GALATSANOS N P. A bayesian framework for image segmentationwith spatially varying mixtures [J]. IEEE Transactions on Image Processing,2010,19(9):2278–2289.
[30] CHEN S F, CAO L L, WANG Y M, LIU J Z, TANG X O. Image segmentation by MAP-MLestimations [J]. IEEE Transactions on Image Processing,2010,19(9):2254–2264.
[31] CHENG L, GONG M L, SCHUURMAN D, CAELLI T. Real-time discriminative backgroundsubtraction [J]. IEEE Transactions on Image Processing,2011,20(5):1401–1414.
[32] CHATZIS S P, TSECHPENAKIS G. The infinite hidden markov random field model [J].IEEE Transactions on Neural Networks,2010,21(6):1004–1014.
[33]陈晓惠,郑晨,段汕,秦前清,胡亦钧.形态小波域多尺度马尔可夫模型在纹理图像分割中的应用[J].中国图象图形学报,2011,16(5):761–766.
[34]刘国英，王爱民，陈荣元，秦前清.基于小波域TS-MRF模型的监督图像分割方法[J].红外与毫米波学报，2011,30(1):91–96.
[35] LEE S H, KOO H I, CHO N I. Image segmentation algorithms based on the machine learningof features [J]. Pattern Recognition Letter,2010,31(14):2325–2336.
[36] LI X, SAHBI H. Superpixel-based object class segmentation using conditional randomfields[C]. In: Proceedings of the IEEE International Conference on Acoustics, Speech, andSignal Processing, Prague, CZECH: IEEE,2011.1101-1104.
[37] Awasthi P, Gagrani A, Ravindran B. Image modeling using tree structured conditional randomfields[C]. In: Proceedings of Artificial Intelligence, Hyderabad, INDIA,2007.2054–2059.
[38] Kapoor A, Winn J. Located hidden random fields: learning discriminative parts for objectdetection [C]. In: Proceedings of the8th European Conference on Computer Vision, Graz,AUSTRIA,2006.302–315.
[39] Winn J, Shotton J. The layout consistent random field for recognizing and segmenting partially occluded objects [C]. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, New York, USA: IEEE,2006.37–44.
[40] Hoiem D, Rother C, Winn J.3D layout CRF for multi-view object class recognition and segmentation[C]. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Minnesota, USA: IEEE,2007.1–8.
[41]高琳,唐鹏,盛鹏.基于前景分割和特征空间自适应选择的视觉目标跟踪[J].控制与决策.2010,25(2):207–212.
[42]耿晓伟,赵杰煜.基于多线索动态融合的交互式图像分割[J].中国图象图形学报.2010,15(1):75–84.
[43] PRINCIPE J C, PUTTHIVIDHYA D, RANGARAJAN A. Data-driven tree-structuredbayesian network for image segmentation [C]. In: Proceedings of the IEEE InternationalConference on Acoustics, Speech and Signal Processing, Kyoto, JAPAN: IEEE,2012.2213–2216.
[44] ALAM F I, GONDRA I. A bayesian network-based tunable image segmentation algorithm forobject recognition [C]. In: Proceedings of the IEEE International Symposium on SignalProcessing and Information Technology, Bilbao, SPAIN: IEEE,2011.11-16.
[45] ZHANG L, JI Q. A bayesian network model for automatic and interactive image segmentation[J]. IEEE Transactions on Image Processing,2011,20(9):2582–2593.
[46]陈聪.非监督贝叶斯彩色图像分割[D].武汉,中南民族大学,2010.
[47]蒋建国,郭艳蓉,郝世杰,詹曙,李鸿, IAN Ross.贝叶斯框架下的非参数估计Graph Cuts分割算法[J].中国图象图形学报.2011,16(6):947–952.
[48] YILMAZ A, JAVED O, SHAN M. Object tracking: a survey [J]. ACM Computing Surveys,2006,38(4):1–45.
[49] YANG H X, SHAO L, ZHENG F, WANG L, SONG Z. Recent advances and trends in visualtracking: a review [J]. Neurocomputing,2011,74:3823–3821.
[50] LI Y, NEVATI R. Key object driven multicategory object recognition, localization andtracking using spatio temporal context[C]. In: Proceedings of the10th European Conferenceon Computer Vision. Marseille, France: Springer,2008.409–422.
[51] STALDER S, GRABNER H, GOOL L. Cascaded confidence filtering for improved trackingby detection[C]. In: Proceedings of the10th European Conference on Computer Vision.Heraklion, Greece: Johnson,2010.369–382.
[52] ROTH P, STERNIG S, GRABNER H, BISCHOF H. Classifier grids for robust adaptive object detection[C]. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Miami Beach, USA: IEEE,2009.2719–2726.
[53]程有龙,李斌,张文聪,庄镇泉.融合先验知识的自适应行人跟踪算法[J].模式识别与人工智能.2009,22(5):704–708.
[54]查于飞,毕笃彦,杨源,董守平,罗宁.基于全局和局部约束直推学习的鲁棒跟踪研究[J].自动化学报.2010,36(8):1084–1090.
[55] DIWALA S, HOIEM D, HAYS J H, EFROS A A, HEBERT M. An empirical study of contextin object detection[C]. In: Proceedings of the IEEE Conference on Computer Vision andPattern Recognition, Miami Beach, USA: IEEE,2009.1271–1278.
[56] LI P H, ZHANG T W, MA B. Unscented kalman filter for visual curve tracking [J]. Imageand Vision Computing,2004,22(2):157–164.
[57] MASKELL S, GORDON N, CLAPP T. A tutorial on particle filters for online nonlinear/nongaussian bayesian tracking [J]. IEEE Transactions on Signal Processing,2002,50(2):174–188.
[58] ZHANG X Q, HU W M, MAYBAN S. A smarter particle filter [C]. In: Proceedings of9thAsian Conference on Computer Vision, xian China,2009.236–246.
[59] KWON J, LEE K. Tracking of a non rigid object via patch based dynamic appearancemodeling and adaptive basin hopping Monte Carlo sampling [C]. In: Proceedings of the IEEEConference on Computer Vision and Pattern Recognition, Miami Beach, USA: IEEE,2009.1208–1215.
[60] KWON J, LEE K. Visual tracking decomposition[C]. In: Proceedings of the23rd IEEEConference on Computer Vision and Pattern Recognition. San Francisco, USA: IEEE,2010.1269–1276.
[61]李远征,卢朝阳,高全学,李静.基于多特征融合的均值迁移粒子滤波跟踪算法[J].电子与信息学报.2010,32(2):411–415.
[62] WEN’ J, GAO X. Incremental learning of weighted tensor subspace for visual tracking[C]. in:Proceedings of the IEEE International Conference on Systems, Man and Cybernetics, SanAntonio, USA: IEEE,2009.3688–3693.
[63] WU Y, CHENG J, WANG J Q, LU H. Real time visual tracking via incremental covariancetensor learning[C]. In: Proceedings of the12th IEEE International Conference on ComputerVision. Kyoto Japan, IEEE,2009.1631–1638.
[64]温静,李洁,高新波.基于增量张量子空间学习的自适应目标跟踪[J].电子学报.2009,37(7):1618–1623.
[65] GRABNER H, BISCHOF H. On-line boosting and vision[C]. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, New York, USA: IEEE2006.260–267.
[66] STALDER S, GRABNER H, GOOL L V. Beyond semi-supervised tracking: tracking shouldbe as simple as detection, but not simpler than recognition[C]. In: Proceedings of the12thIEEE International Conference on Computer Vision, Kyoto Japan: IEEE,2009.1409–1416.
[67] ZEISL B, LEISTNER C, SAFFARI A, BISCHOF H. On-line semi-supervisedmultiple-instance boosting[C]. In: Proceedings of the IEEE Conference on Computer Visionand Pattern Recognition. San Francisco, USA: IEEE,2010.1879–1886.
[68] KIM T K, WOODLEY T, STENGER B, CIPOLLA R. Online multiple classifier boosting forobject Tracking[C]. In: Proceedings of the IEEE Workshop on Computer Vision and PatternRecognition, San Francisco, USA: IEEE,2010.1–6.
[69]贾静平,张飞,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.
[70] KSCHISCHANG F R, FREY B J, LOELIGER H A. Factor graphs and the sum-productalgorithm [J]. IEEE Transactions on Information Theory,2001,47(2):498–519.
[71] BISHOP C M. Pattern recognition and machine learning [M]. Springer,2006.
[72]张林,刘辉. Dirichlet过程混合模型的聚类算法[J].中国矿业大学学报,2012,41(1):159–163.
[73]周建英王飞跃曾大军.分层Dirichlet过程及其应用综述[J].自动化学报2011,37(4):389–407.
[74] FERGUSON T S. A Bayesian analysis of some nonparametric problems [J]. The Annals ofstatistics,1973,1(2):209–230.
[75] SUDDERTH E B. Graphical models for visual object recognition and tracking [D]. USA:Department of Electrical Engineering and Computer Science, Massachusetts Institute ofTechnology,2006.
[76] PITMAN J. Combinatorial stochastic processes [R]. Lecture Notes in Mathematics1875.Berlin: Springer-Verlag,2006.
[77] THE Y W, JORDAN M I, BEAL M J, BLEI D M. Hierarchical Dirichlet processes [J].Journal of the American Statistical Association,2006,101(476):1566–1581.
[78] BEAL M J, GHAHRAMANI Z, RASMUSSEN C E. The infinite hidden Markov model [C].In: Proceedings of the Advances in Neural Information Processing Systems, Vancouver,Canada: MIT Press,2001.577–584.
[79] BOYKOV Y, FUNKA-LEA G. Graph cuts and efficient N-D image segmentation [J].International Journal of Computer Vision,2006,70(2):109–131.
[80] KOLMOGOROV V, ZABIH R. What energy functions can be minimized via graph cuts?[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2004,26(2):147–159.
[81] MALIK J, PERONA P. Scale-space and edge detection using anisotropic diffusion [J]. IEEETransactions on Pattern Analysis and Machine Intelligence,1990,12(7):629–639.
[82] DENG Y H, WANG Y Y, SHEN Y Z. Speckle reduction of ultrasound images based onRayleigh-trimmed anisotropic diffusion filter [J]. Pattern Recognition Letters,201132(1):1516–1525.
[83] JANEV M, PILIPOVIC S, ATANAACKOVIC T, OBRADOVIC R, RALEVIC N. Fullyfractional anisotropic diffusion for image denoising [J]. Mathematical and ComputerModelling,2011,54(1):729–741.
[84] CHAO Shin-Min, TSAI Du-Ming. An improved anisotropic diffusion model for detail-andedge-preserving smoothing [J]. Pattern Recognition Letters,2010,31:2012–2023.
[85] SUN Jian, XU Zong-Ben. Scale selection for anisotropic diffusion filter by Markov randomfield model [J]. Pattern Recognition,2010,43:2630–2645.
[86] GRASMAIR M，LENZEN F. Anisotropic total variation filtering [J]. Appl Math Optim,2010,62:323–339.
[87] ROI M.R, MARIA C.C, JULIO M.H. Accurate implementation of anisotropic diffusion in thehypercube [J]. IEEE Geoscience and Remote Sensing Letters,2010,7(4):870–874.
[88]崔丽洁,郑江滨,李秀秀.基于SVD背景抑制和粒子滤波的弱小目标检测[J].计算机应用研究,2011,28(4):1553–1555.
[89] ZHANG W G, LIU F, JIAO L C，HOU B A，WANG S，SHANG R H. SAR image despecklingusing edge detection and feature clustering in bandelet domain [J]. IEEE Geoscience andRemote Sensing Letters,2010,7(1):131–135.
[90]陈卿，胡绍湘.基于边缘扩散的医学图像非线性去噪算法[J].计算机应用研究,2011,28(8):3178–3180.
[91] LOPEZ M C, DE B, BUSTINCE H. Generating fuzzy edge images from gradient magnitudes[J]. Compution Vision and Image Understanding,2011,115(11):1571–1580.
[92] PUVANATHASAN P, BIZHEVAL K. Interval type-II fuzzy anisotropic diffusion algorithmfor speckle noise reduction in optical coherence tomography images [J]. Optics Express,2009,17(2):733–746.
[93] DEMIRCI R. Fuzzy adaptive anisotropic filter for medical images [J]. Expert Systems,2010,27(3):219–229.
[94] LOWE D G. Distinctive image features from scale-invariant keypoints [J]. InternationalJournal of Computer Vision,2004,60(2):91–110.
[95] SOYEL H, DEMIREL H. Facial expression recognition based on discriminative scaleinvariant feature transformation [J]. Electronics Letters,2010,46(5):343–u4863.
[96] PSYLLOS A P, ANAGNOSTOPOULOS C-N E, KAYAFAS E. Vehicle logo recognitionusing a sift-Based enhanced matching scheme [J]. IEEE Transactions on IntelligentTransportation Systems,2010,11(2):322–328.
[97] SEDAGHAT A, MOKHTARZADE M, EBADI H. Uniform robust scale-invariant featurematching for optical remote sensing images [J]. IEEE Transactions Geoscience and RemoteSensing,2011,49(11):4516–4527.
[98] RACHEL LO T-W, PAUL SIEBERT J. Local feature extraction and matching on rangeimages:2.5D SIFT [J]. Computer Vision and Image Understanding,2009,113:1235–1250.
[99] LIU X Z, TIAN Z, DING M G. A novel adaptive weights proximity matrix for imageregistration based on R-SIFT [J]. International Journal of Electronics and Communications,2011,65(12):1040–1049.
[100] ZHANG X H, WANG H X, HONG M J, YANG D, LOVELLC B C. Robust image cornerdetection based on scale evolution difference of planar curves [J]. Pattern Recognition Letters,2009,30(4):449–455.
[101] BAY H, ESS A, TUYTELAARS T, VAN G. Speeded-up robust features (SURF)[J].Computer Vision and Image Understanding,2008,110(3):346–359.
[102]吴莉,梅雪,林锦国.基于Contourlet变换和KFD的相似目标特征提取[J].计算机工程与设计,2011,32(1):240–243.
[103]孙鑫,谢元旦,任地成.基于小波变换和子图的图像配准方法[J].计算机工程与设计,2010,31(21):4653–4654.
[104] CREMERS D, ROUSSON M, DERICHE R. A review of statistical approaches to level setsegmentation: integrating color, texture, motion and shape [J]. International Journal ofComputer Vision,2007,72(2):195–215.
[105] CREMERS D, OSHER S J, SOATTO S. Kernel density estimation and intrinsic alignment forshape priors in level set segmentation [J]. International Journal of Computer Vision,2006,69(3):335–351.
[106] CHAN T, ZHU WEI. Level set based shape prior segmentation[C]. In: Proceedings of theIEEE Conference on Computer Vision and Pattern Recognition, San Diego, USA: IEEE,2005.2:1164–1170.
[107]吴晓红,罗代升,王正勇，唐伟力，赵文斌.基于边缘流的多尺度水平集砾岩图像分割[J].四川大学学报(工程科学版),2008,40(1):133–137.
[108] FUNDANA K, OVERGAAD NC, HEYDEN A. Variational segmentation of imagesequences using region-based active contours and deformable shape Priors [J]. InternationalJournal of Computer Vision,2008,80:289–299.
[109] ALESSANDRINI M, DIETENBECK T, BASSET O, FRIBOULET D, BERNARD O. Usinga geometric formulation of annular-like shape priors for constraining variational level-sets[C].In: Proceedings of the IEEE International Conference on Image Processing, Hong Kong,CHINA: IEEE,2010.669–672.
[110] GRADY L, ALVINO C. The piecewise smooth mumford-shah functional on an arbitrarygraph [J]. IEEE Transaction on Image Processing,2009,18(11):2547–2561.
[111] EI-ZEHIRY N, SAHOO P, ELMAGHRABY A. Combinatorial optimization of the piecewiseconstant Mumford-Shah functional with application to scalar/vector valued and volumetricimage segmentation [J]. Image and Vision Computing,2011,29:365–381.
[112] SZELISKI R, ZABIH R, SCHARSTEIN D, VEKSLER O, KOLMOGOROV V. Acomparative study of energy minimization methods for Markov random fields withsmoothness-based priors [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2008,30(6):1068–1080.
[113]周激流,詹晓倩,何坤,刘昶.基于统计估计的图像边缘检测[J].沈阳工业大学学报),2010,32(6):665–670.
[114] CHITTAJALLU D R, SHAH SK, KAKADIARIS I A. A shape-driven MRF model for thesegmentation of organs in medical images [C]. In: Proceedings of the IEEE Conference onComputer Vision and Pattern Recognition. San Francisco, USA: IEEE,2010,3233–3240.
[115] DO M N, VETTERLI M. The contourlet transform: an efficient directional multiresolutionimage representation [J]. IEEE Transactions Image on Processing,2005,14(12):2091–2106.
[116] DA CUNHA A.L, ZHOU Jan-Ping, DO M.N. The nonsubsampled contourlet transform:theory, design, and applications [J]. IEEE Transaction on Image Processing,2006,15(10):3089–3101.
[117] http://vision.ece.ucsb.edu/download.html.
[118] BHUTADA G.G, ANAND S.G, SAXENA S.C. Edge preserved image enhancement usingadaptive fusion of images denoised by wavelet and curvelet transform [J]. Digital SignalProcessing,2011,21:118–130.
[119] MIKOLAJCZYK K, SCHMID C. A performance evaluation of local descriptors [J]. IEEETransactions on Pattern Analysis and Machine Intelligence,2005,27(10):1615–1630.
[120] BESBES A, KOMODAKIS N, LANGS G, PARAGIOS N. Shape priors and discrete MRFsfor knowledge-based segmentation[C]. In: Proceedings of the IEEE Conference on ComputerVision and Pattern Recognition, Miami Beach, USA: IEEE,2009.4:1295–1302.
[121] THIRUVENKADAM S R, CHAN T F, HONG B W. Segmentation under occlusions usingselective shape prior [C]. International Conference on Scale Space and Variational Methods inComputer Vision, Ischia, ITALY:2007.191–202.
[122] CIESIELSKI K C, UDUPA J K. A framework for comparing different image segmentationmethods and its use in studying equivalences between level set and fuzzy connectednessframeworks [J]. Computer Vision and Image Understanding,2011,115:721–734.
[123] CASELLES V, KIMMEL R, SAPIRO G. Geodesic active contours [J]. International Journalof Computer Vision,1997,22(1):61–79.
[124] LI Chun-Ming, KAO Chiu-Yen, GORE J C, DING Z H. Minimization of region-scalablefitting energy for image segmentation [J]. IEEE Transaction on Image Processing,2008,17(10):1940–1949.
[125] LANKTON S, TANNENBAUM A. Localizing region-based active contours [J]. IEEETransaction on Image Processing,2008,17(11):2029–2039.
[126] ZHANG Kai-Hua, SONG Hui-Hui, ZHANG Lei. Active contours driven by local imagefitting energy [J]. Pattern Recognition，2010,43:1199–1206.
[127]王利,陈允杰,韦志辉,夏德深,王平安.克服灰度不均匀性的脑MR图像分割模型[J].计算机辅助设计与图形学学报，2009,21(11):1624–1631.
[128]刘剑磊,冯大政.一种全局最优的非匀质图像分割方法[J].西安电子科技大学学报(自然科学版),2011,38(2):66–71.
[129] LIU Shi-gang, PENG Ya-li. A local region-based Chan–Vese model for image segmentation[J]. Pattern Recognition，2012,45:2769–2779.
[130] AVIDAN S. Ensemble tracking [J]. IEEE Transactions on Pattern Analysis and MachineIntelligence,2007,29(2):261–271.
[131]仝小敏,张艳宁,杨涛.基于增量子空间自适应决策的目标跟踪[J].自动化学报,2011,37(12):1483–1494.
[132] PARK D W, KWON J, LEE K M. Robust visual tracking using autoregressive hidden Markovmodel[C]. In: Proceedings of the IEEE Conference on Computer Vision and PatternRecognition. Providence RI, USA: IEEE,2012.1964–1971.
[133]王书朋,姬红兵.用于目标跟踪的自适应粒子滤波算法[J].系统仿真学报,2010,22(3):630–633.
[134] AHMED A, XING E. Dynamic non-parametric mixture models and the recurrent Chineserestaurant process: with applications to evolutionary clustering[C]. In: Proceedings of theSIAM international conference on data mining. Atlanta, USA: SIAM,2008.219–230.
[135] XU T B, ZHANG Z F, YU P S, LONG Bo. Dirichlet process based evolutionary clustering[C].In: Proceedings of8th IEEE International Conference on Data Ming. Pisa, Italy: IEEE,2008.648–657.
[136] SETHURAMAN J. A constructive definition of Dirichlet priors [J]. Statistica Sinica,1994,4:639–650.
[2] LANKTON M S. localized statistical model in computer vision [D]. USA. Georgia Instituteof Technology,2009.
[3] Koller D, Friedman N. Probabilistic Graphical Models Principles and Technique [M].Cambridge, MA: MIT Press,2009.
[4]高琳,唐鹏,盛鹏.基于概率图模型目标建模的视觉跟踪算法[J].光电子·激光,2010,21(1):124–129.
[5] LI J, ALLINSON N M. A comprehensive review of current local features for computer vision[J]. Neurocomputing,2008,71:1771–1787.
[6] SCHELTEN K, ROTH S. Connecting non-quadratic variational models and MRFs [C]. In:Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, ColoradoSprings, USA: IEEE,2011.2641–2641.
[7] ZHANG L, JI Q. Image segmentation with a unified graphical model [J]. IEEE Transactionson Pattern Analysis and Machine Intelligence,2010,32(8):1406–1425.
[8] WU T, BAE M H, ZHANG M, PAN R, BADEA A. A prior feature SVM-MRF basedmethod for mouse brain segmentation [J]. Neuroimage,2012,59(3):2298–2306.
[9]卢易苏,陈武凡.非参数图像分割算法的研究与改进[J].计算机工程,2012,38(7):179–181.
[10] GRABNER H, LEISTNER C, BISCHOF H. Semi-supervised online boosting for robust tracking [C]. In: Proceedings of the10th European Conference on Computer Vision. Marseille, France:2008.234–247.
[11] BABENKO B, YANG M H, BELONGIE S. Robust object tracking with online multiple instance learning[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2011,33(8):1619–1632.
[12] SANTNER J, LEISTNER C, SAFFARI A, POCK T, BISCHOF H. PROST parallel robustonline simple tracking [C]. In: Proceedings of the IEEE Conference on Computer Vision andPattern Recognition. San Francisco, USA: IEEE,2010.723–730.
[13] PAN S J, YANG Q. A survey on transfer learning [J]. IEEE Transactions on Knowledge and Data Engineering,2010,22(10):1345–1359.
[14] WHITTAKER J. Graphical models in applied multivariate statistics [M]. New York: JohnWiley and Sons lnc,1990.
[15] JORDAN M I, GHAHRAMANI Z, JAAKKOLA T, SAUL L K. An introduction to variationalmethods for graphical models [J]. Machine Learning,1999,37(2):183–233.
[16] GILKS W R, RICHARDSON S, SPIEGELHALTER D J. Markov chain Monte Carlo in practice [M]. London: Chapman and Hall,1996.
[17] PEARL J. Probabilistic reasoning in intelligent systems: networks of plausible inference [M].San Mateo, California, USA: Morgan Kaufmann Publishers Inc,1988.
[18]朱允刚.贝叶斯网学习中若干问题研究及其在信息融合中的应用[D].长春:吉林大学,2012.
[19]王相海,陈明莹,方玲玲.概率图模型及其图像与视频应用研究[J].中国图象图形学报,2009,14(9):1712–1720.
[20]盛骤,谢式千,潘承毅.概率论与数理统计[M].浙江:高等教育出版社,2003.
[21] PEARL J. Bayes and Markov networks: a comparison of two graphical representations of probabilistic knowledge [R].860024(R46), Los Angeles, California, USA: University of California at Los Angeles,1986.
[22]刘李漫.基于变分模型和图割优化的图像与视频目标分割方法研究[D].武汉:华中科技大学,2012.
[23] COMANICIU D, MEER P. Mean shift: a robust approach toward feature space analysis [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(5):603–619.
[24] KASS M, WITKIN A, TERZOPOULOS D. Snakes: active contour models [J]. International Journal of Computer Visio,1988,1:321–331.
[25] SHI J, MALIK J. Normalized cuts and image segmentation [J]. IEEE Transactions on PatternAnalysis and Machine Intelligence,2000,22(8):888–905.
[26] CHAN T, VESE L. Active contours without edges [J]. IEEE Transactions on Image Processing,2001,10(2):266–277.
[27] LI STAN Z. Markov random field modeling in image analysis [M]. Third edition. Springer,2009.
[28] Lafferty J, McCallum A, Pereira F. Conditional random fields: probabilistic models forsegmention and labeling sequence data[C]. In: Proceedings of Machine Learning, Freiburg,GERMANY,2001.282–289.
[29] NIKOU C, LIKAS A C, GALATSANOS N P. A bayesian framework for image segmentationwith spatially varying mixtures [J]. IEEE Transactions on Image Processing,2010,19(9):2278–2289.
[30] CHEN S F, CAO L L, WANG Y M, LIU J Z, TANG X O. Image segmentation by MAP-MLestimations [J]. IEEE Transactions on Image Processing,2010,19(9):2254–2264.
[31] CHENG L, GONG M L, SCHUURMAN D, CAELLI T. Real-time discriminative backgroundsubtraction [J]. IEEE Transactions on Image Processing,2011,20(5):1401–1414.
[32] CHATZIS S P, TSECHPENAKIS G. The infinite hidden markov random field model [J].IEEE Transactions on Neural Networks,2010,21(6):1004–1014.
[33]陈晓惠,郑晨,段汕,秦前清,胡亦钧.形态小波域多尺度马尔可夫模型在纹理图像分割中的应用[J].中国图象图形学报,2011,16(5):761–766.
[34]刘国英，王爱民，陈荣元，秦前清.基于小波域TS-MRF模型的监督图像分割方法[J].红外与毫米波学报，2011,30(1):91–96.
[35] LEE S H, KOO H I, CHO N I. Image segmentation algorithms based on the machine learningof features [J]. Pattern Recognition Letter,2010,31(14):2325–2336.
[36] LI X, SAHBI H. Superpixel-based object class segmentation using conditional randomfields[C]. In: Proceedings of the IEEE International Conference on Acoustics, Speech, andSignal Processing, Prague, CZECH: IEEE,2011.1101-1104.
[37] Awasthi P, Gagrani A, Ravindran B. Image modeling using tree structured conditional randomfields[C]. In: Proceedings of Artificial Intelligence, Hyderabad, INDIA,2007.2054–2059.
[38] Kapoor A, Winn J. Located hidden random fields: learning discriminative parts for objectdetection [C]. In: Proceedings of the8th European Conference on Computer Vision, Graz,AUSTRIA,2006.302–315.
[39] Winn J, Shotton J. The layout consistent random field for recognizing and segmenting partially occluded objects [C]. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, New York, USA: IEEE,2006.37–44.
[40] Hoiem D, Rother C, Winn J.3D layout CRF for multi-view object class recognition and segmentation[C]. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Minnesota, USA: IEEE,2007.1–8.
[41]高琳,唐鹏,盛鹏.基于前景分割和特征空间自适应选择的视觉目标跟踪[J].控制与决策.2010,25(2):207–212.
[42]耿晓伟,赵杰煜.基于多线索动态融合的交互式图像分割[J].中国图象图形学报.2010,15(1):75–84.
[43] PRINCIPE J C, PUTTHIVIDHYA D, RANGARAJAN A. Data-driven tree-structuredbayesian network for image segmentation [C]. In: Proceedings of the IEEE InternationalConference on Acoustics, Speech and Signal Processing, Kyoto, JAPAN: IEEE,2012.2213–2216.
[44] ALAM F I, GONDRA I. A bayesian network-based tunable image segmentation algorithm forobject recognition [C]. In: Proceedings of the IEEE International Symposium on SignalProcessing and Information Technology, Bilbao, SPAIN: IEEE,2011.11-16.
[45] ZHANG L, JI Q. A bayesian network model for automatic and interactive image segmentation[J]. IEEE Transactions on Image Processing,2011,20(9):2582–2593.
[46]陈聪.非监督贝叶斯彩色图像分割[D].武汉,中南民族大学,2010.
[47]蒋建国,郭艳蓉,郝世杰,詹曙,李鸿, IAN Ross.贝叶斯框架下的非参数估计Graph Cuts分割算法[J].中国图象图形学报.2011,16(6):947–952.
[48] YILMAZ A, JAVED O, SHAN M. Object tracking: a survey [J]. ACM Computing Surveys,2006,38(4):1–45.
[49] YANG H X, SHAO L, ZHENG F, WANG L, SONG Z. Recent advances and trends in visualtracking: a review [J]. Neurocomputing,2011,74:3823–3821.
[50] LI Y, NEVATI R. Key object driven multicategory object recognition, localization andtracking using spatio temporal context[C]. In: Proceedings of the10th European Conferenceon Computer Vision. Marseille, France: Springer,2008.409–422.
[51] STALDER S, GRABNER H, GOOL L. Cascaded confidence filtering for improved trackingby detection[C]. In: Proceedings of the10th European Conference on Computer Vision.Heraklion, Greece: Johnson,2010.369–382.
[52] ROTH P, STERNIG S, GRABNER H, BISCHOF H. Classifier grids for robust adaptive object detection[C]. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Miami Beach, USA: IEEE,2009.2719–2726.
[53]程有龙,李斌,张文聪,庄镇泉.融合先验知识的自适应行人跟踪算法[J].模式识别与人工智能.2009,22(5):704–708.
[54]查于飞,毕笃彦,杨源,董守平,罗宁.基于全局和局部约束直推学习的鲁棒跟踪研究[J].自动化学报.2010,36(8):1084–1090.
[55] DIWALA S, HOIEM D, HAYS J H, EFROS A A, HEBERT M. An empirical study of contextin object detection[C]. In: Proceedings of the IEEE Conference on Computer Vision andPattern Recognition, Miami Beach, USA: IEEE,2009.1271–1278.
[56] LI P H, ZHANG T W, MA B. Unscented kalman filter for visual curve tracking [J]. Imageand Vision Computing,2004,22(2):157–164.
[57] MASKELL S, GORDON N, CLAPP T. A tutorial on particle filters for online nonlinear/nongaussian bayesian tracking [J]. IEEE Transactions on Signal Processing,2002,50(2):174–188.
[58] ZHANG X Q, HU W M, MAYBAN S. A smarter particle filter [C]. In: Proceedings of9thAsian Conference on Computer Vision, xian China,2009.236–246.
[59] KWON J, LEE K. Tracking of a non rigid object via patch based dynamic appearancemodeling and adaptive basin hopping Monte Carlo sampling [C]. In: Proceedings of the IEEEConference on Computer Vision and Pattern Recognition, Miami Beach, USA: IEEE,2009.1208–1215.
[60] KWON J, LEE K. Visual tracking decomposition[C]. In: Proceedings of the23rd IEEEConference on Computer Vision and Pattern Recognition. San Francisco, USA: IEEE,2010.1269–1276.
[61]李远征,卢朝阳,高全学,李静.基于多特征融合的均值迁移粒子滤波跟踪算法[J].电子与信息学报.2010,32(2):411–415.
[62] WEN’ J, GAO X. Incremental learning of weighted tensor subspace for visual tracking[C]. in:Proceedings of the IEEE International Conference on Systems, Man and Cybernetics, SanAntonio, USA: IEEE,2009.3688–3693.
[63] WU Y, CHENG J, WANG J Q, LU H. Real time visual tracking via incremental covariancetensor learning[C]. In: Proceedings of the12th IEEE International Conference on ComputerVision. Kyoto Japan, IEEE,2009.1631–1638.
[64]温静,李洁,高新波.基于增量张量子空间学习的自适应目标跟踪[J].电子学报.2009,37(7):1618–1623.
[65] GRABNER H, BISCHOF H. On-line boosting and vision[C]. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, New York, USA: IEEE2006.260–267.
[66] STALDER S, GRABNER H, GOOL L V. Beyond semi-supervised tracking: tracking shouldbe as simple as detection, but not simpler than recognition[C]. In: Proceedings of the12thIEEE International Conference on Computer Vision, Kyoto Japan: IEEE,2009.1409–1416.
[67] ZEISL B, LEISTNER C, SAFFARI A, BISCHOF H. On-line semi-supervisedmultiple-instance boosting[C]. In: Proceedings of the IEEE Conference on Computer Visionand Pattern Recognition. San Francisco, USA: IEEE,2010.1879–1886.
[68] KIM T K, WOODLEY T, STENGER B, CIPOLLA R. Online multiple classifier boosting forobject Tracking[C]. In: Proceedings of the IEEE Workshop on Computer Vision and PatternRecognition, San Francisco, USA: IEEE,2010.1–6.
[69]贾静平,张飞,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.
[70] KSCHISCHANG F R, FREY B J, LOELIGER H A. Factor graphs and the sum-productalgorithm [J]. IEEE Transactions on Information Theory,2001,47(2):498–519.
[71] BISHOP C M. Pattern recognition and machine learning [M]. Springer,2006.
[72]张林,刘辉. Dirichlet过程混合模型的聚类算法[J].中国矿业大学学报,2012,41(1):159–163.
[73]周建英王飞跃曾大军.分层Dirichlet过程及其应用综述[J].自动化学报2011,37(4):389–407.
[74] FERGUSON T S. A Bayesian analysis of some nonparametric problems [J]. The Annals ofstatistics,1973,1(2):209–230.
[75] SUDDERTH E B. Graphical models for visual object recognition and tracking [D]. USA:Department of Electrical Engineering and Computer Science, Massachusetts Institute ofTechnology,2006.
[76] PITMAN J. Combinatorial stochastic processes [R]. Lecture Notes in Mathematics1875.Berlin: Springer-Verlag,2006.
[77] THE Y W, JORDAN M I, BEAL M J, BLEI D M. Hierarchical Dirichlet processes [J].Journal of the American Statistical Association,2006,101(476):1566–1581.
[78] BEAL M J, GHAHRAMANI Z, RASMUSSEN C E. The infinite hidden Markov model [C].In: Proceedings of the Advances in Neural Information Processing Systems, Vancouver,Canada: MIT Press,2001.577–584.
[79] BOYKOV Y, FUNKA-LEA G. Graph cuts and efficient N-D image segmentation [J].International Journal of Computer Vision,2006,70(2):109–131.
[80] KOLMOGOROV V, ZABIH R. What energy functions can be minimized via graph cuts?[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2004,26(2):147–159.
[81] MALIK J, PERONA P. Scale-space and edge detection using anisotropic diffusion [J]. IEEETransactions on Pattern Analysis and Machine Intelligence,1990,12(7):629–639.
[82] DENG Y H, WANG Y Y, SHEN Y Z. Speckle reduction of ultrasound images based onRayleigh-trimmed anisotropic diffusion filter [J]. Pattern Recognition Letters,201132(1):1516–1525.
[83] JANEV M, PILIPOVIC S, ATANAACKOVIC T, OBRADOVIC R, RALEVIC N. Fullyfractional anisotropic diffusion for image denoising [J]. Mathematical and ComputerModelling,2011,54(1):729–741.
[84] CHAO Shin-Min, TSAI Du-Ming. An improved anisotropic diffusion model for detail-andedge-preserving smoothing [J]. Pattern Recognition Letters,2010,31:2012–2023.
[85] SUN Jian, XU Zong-Ben. Scale selection for anisotropic diffusion filter by Markov randomfield model [J]. Pattern Recognition,2010,43:2630–2645.
[86] GRASMAIR M，LENZEN F. Anisotropic total variation filtering [J]. Appl Math Optim,2010,62:323–339.
[87] ROI M.R, MARIA C.C, JULIO M.H. Accurate implementation of anisotropic diffusion in thehypercube [J]. IEEE Geoscience and Remote Sensing Letters,2010,7(4):870–874.
[88]崔丽洁,郑江滨,李秀秀.基于SVD背景抑制和粒子滤波的弱小目标检测[J].计算机应用研究,2011,28(4):1553–1555.
[89] ZHANG W G, LIU F, JIAO L C，HOU B A，WANG S，SHANG R H. SAR image despecklingusing edge detection and feature clustering in bandelet domain [J]. IEEE Geoscience andRemote Sensing Letters,2010,7(1):131–135.
[90]陈卿，胡绍湘.基于边缘扩散的医学图像非线性去噪算法[J].计算机应用研究,2011,28(8):3178–3180.
[91] LOPEZ M C, DE B, BUSTINCE H. Generating fuzzy edge images from gradient magnitudes[J]. Compution Vision and Image Understanding,2011,115(11):1571–1580.
[92] PUVANATHASAN P, BIZHEVAL K. Interval type-II fuzzy anisotropic diffusion algorithmfor speckle noise reduction in optical coherence tomography images [J]. Optics Express,2009,17(2):733–746.
[93] DEMIRCI R. Fuzzy adaptive anisotropic filter for medical images [J]. Expert Systems,2010,27(3):219–229.
[94] LOWE D G. Distinctive image features from scale-invariant keypoints [J]. InternationalJournal of Computer Vision,2004,60(2):91–110.
[95] SOYEL H, DEMIREL H. Facial expression recognition based on discriminative scaleinvariant feature transformation [J]. Electronics Letters,2010,46(5):343–u4863.
[96] PSYLLOS A P, ANAGNOSTOPOULOS C-N E, KAYAFAS E. Vehicle logo recognitionusing a sift-Based enhanced matching scheme [J]. IEEE Transactions on IntelligentTransportation Systems,2010,11(2):322–328.
[97] SEDAGHAT A, MOKHTARZADE M, EBADI H. Uniform robust scale-invariant featurematching for optical remote sensing images [J]. IEEE Transactions Geoscience and RemoteSensing,2011,49(11):4516–4527.
[98] RACHEL LO T-W, PAUL SIEBERT J. Local feature extraction and matching on rangeimages:2.5D SIFT [J]. Computer Vision and Image Understanding,2009,113:1235–1250.
[99] LIU X Z, TIAN Z, DING M G. A novel adaptive weights proximity matrix for imageregistration based on R-SIFT [J]. International Journal of Electronics and Communications,2011,65(12):1040–1049.
[100] ZHANG X H, WANG H X, HONG M J, YANG D, LOVELLC B C. Robust image cornerdetection based on scale evolution difference of planar curves [J]. Pattern Recognition Letters,2009,30(4):449–455.
[101] BAY H, ESS A, TUYTELAARS T, VAN G. Speeded-up robust features (SURF)[J].Computer Vision and Image Understanding,2008,110(3):346–359.
[102]吴莉,梅雪,林锦国.基于Contourlet变换和KFD的相似目标特征提取[J].计算机工程与设计,2011,32(1):240–243.
[103]孙鑫,谢元旦,任地成.基于小波变换和子图的图像配准方法[J].计算机工程与设计,2010,31(21):4653–4654.
[104] CREMERS D, ROUSSON M, DERICHE R. A review of statistical approaches to level setsegmentation: integrating color, texture, motion and shape [J]. International Journal ofComputer Vision,2007,72(2):195–215.
[105] CREMERS D, OSHER S J, SOATTO S. Kernel density estimation and intrinsic alignment forshape priors in level set segmentation [J]. International Journal of Computer Vision,2006,69(3):335–351.
[106] CHAN T, ZHU WEI. Level set based shape prior segmentation[C]. In: Proceedings of theIEEE Conference on Computer Vision and Pattern Recognition, San Diego, USA: IEEE,2005.2:1164–1170.
[107]吴晓红,罗代升,王正勇，唐伟力，赵文斌.基于边缘流的多尺度水平集砾岩图像分割[J].四川大学学报(工程科学版),2008,40(1):133–137.
[108] FUNDANA K, OVERGAAD NC, HEYDEN A. Variational segmentation of imagesequences using region-based active contours and deformable shape Priors [J]. InternationalJournal of Computer Vision,2008,80:289–299.
[109] ALESSANDRINI M, DIETENBECK T, BASSET O, FRIBOULET D, BERNARD O. Usinga geometric formulation of annular-like shape priors for constraining variational level-sets[C].In: Proceedings of the IEEE International Conference on Image Processing, Hong Kong,CHINA: IEEE,2010.669–672.
[110] GRADY L, ALVINO C. The piecewise smooth mumford-shah functional on an arbitrarygraph [J]. IEEE Transaction on Image Processing,2009,18(11):2547–2561.
[111] EI-ZEHIRY N, SAHOO P, ELMAGHRABY A. Combinatorial optimization of the piecewiseconstant Mumford-Shah functional with application to scalar/vector valued and volumetricimage segmentation [J]. Image and Vision Computing,2011,29:365–381.
[112] SZELISKI R, ZABIH R, SCHARSTEIN D, VEKSLER O, KOLMOGOROV V. Acomparative study of energy minimization methods for Markov random fields withsmoothness-based priors [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2008,30(6):1068–1080.
[113]周激流,詹晓倩,何坤,刘昶.基于统计估计的图像边缘检测[J].沈阳工业大学学报),2010,32(6):665–670.
[114] CHITTAJALLU D R, SHAH SK, KAKADIARIS I A. A shape-driven MRF model for thesegmentation of organs in medical images [C]. In: Proceedings of the IEEE Conference onComputer Vision and Pattern Recognition. San Francisco, USA: IEEE,2010,3233–3240.
[115] DO M N, VETTERLI M. The contourlet transform: an efficient directional multiresolutionimage representation [J]. IEEE Transactions Image on Processing,2005,14(12):2091–2106.
[116] DA CUNHA A.L, ZHOU Jan-Ping, DO M.N. The nonsubsampled contourlet transform:theory, design, and applications [J]. IEEE Transaction on Image Processing,2006,15(10):3089–3101.
[117] http://vision.ece.ucsb.edu/download.html.
[118] BHUTADA G.G, ANAND S.G, SAXENA S.C. Edge preserved image enhancement usingadaptive fusion of images denoised by wavelet and curvelet transform [J]. Digital SignalProcessing,2011,21:118–130.
[119] MIKOLAJCZYK K, SCHMID C. A performance evaluation of local descriptors [J]. IEEETransactions on Pattern Analysis and Machine Intelligence,2005,27(10):1615–1630.
[120] BESBES A, KOMODAKIS N, LANGS G, PARAGIOS N. Shape priors and discrete MRFsfor knowledge-based segmentation[C]. In: Proceedings of the IEEE Conference on ComputerVision and Pattern Recognition, Miami Beach, USA: IEEE,2009.4:1295–1302.
[121] THIRUVENKADAM S R, CHAN T F, HONG B W. Segmentation under occlusions usingselective shape prior [C]. International Conference on Scale Space and Variational Methods inComputer Vision, Ischia, ITALY:2007.191–202.
[122] CIESIELSKI K C, UDUPA J K. A framework for comparing different image segmentationmethods and its use in studying equivalences between level set and fuzzy connectednessframeworks [J]. Computer Vision and Image Understanding,2011,115:721–734.
[123] CASELLES V, KIMMEL R, SAPIRO G. Geodesic active contours [J]. International Journalof Computer Vision,1997,22(1):61–79.
[124] LI Chun-Ming, KAO Chiu-Yen, GORE J C, DING Z H. Minimization of region-scalablefitting energy for image segmentation [J]. IEEE Transaction on Image Processing,2008,17(10):1940–1949.
[125] LANKTON S, TANNENBAUM A. Localizing region-based active contours [J]. IEEETransaction on Image Processing,2008,17(11):2029–2039.
[126] ZHANG Kai-Hua, SONG Hui-Hui, ZHANG Lei. Active contours driven by local imagefitting energy [J]. Pattern Recognition，2010,43:1199–1206.
[127]王利,陈允杰,韦志辉,夏德深,王平安.克服灰度不均匀性的脑MR图像分割模型[J].计算机辅助设计与图形学学报，2009,21(11):1624–1631.
[128]刘剑磊,冯大政.一种全局最优的非匀质图像分割方法[J].西安电子科技大学学报(自然科学版),2011,38(2):66–71.
[129] LIU Shi-gang, PENG Ya-li. A local region-based Chan–Vese model for image segmentation[J]. Pattern Recognition，2012,45:2769–2779.
[130] AVIDAN S. Ensemble tracking [J]. IEEE Transactions on Pattern Analysis and MachineIntelligence,2007,29(2):261–271.
[131]仝小敏,张艳宁,杨涛.基于增量子空间自适应决策的目标跟踪[J].自动化学报,2011,37(12):1483–1494.
[132] PARK D W, KWON J, LEE K M. Robust visual tracking using autoregressive hidden Markovmodel[C]. In: Proceedings of the IEEE Conference on Computer Vision and PatternRecognition. Providence RI, USA: IEEE,2012.1964–1971.
[133]王书朋,姬红兵.用于目标跟踪的自适应粒子滤波算法[J].系统仿真学报,2010,22(3):630–633.
[134] AHMED A, XING E. Dynamic non-parametric mixture models and the recurrent Chineserestaurant process: with applications to evolutionary clustering[C]. In: Proceedings of theSIAM international conference on data mining. Atlanta, USA: SIAM,2008.219–230.
[135] XU T B, ZHANG Z F, YU P S, LONG Bo. Dirichlet process based evolutionary clustering[C].In: Proceedings of8th IEEE International Conference on Data Ming. Pisa, Italy: IEEE,2008.648–657.
[136] SETHURAMAN J. A constructive definition of Dirichlet priors [J]. Statistica Sinica,1994,4:639–650.