视频图像序列中运动目标检测和跟踪的研究
|
摘要
运动目标的检测与跟踪是计算机视觉中的关键课题。许多领域都要用到该技术,如人机交互、医学诊断、智能视频监控以及智能机器人等。研究中需要运用到图像处理、模式识别、人工智能等很多学科知识。
本文对运动目标的检测与跟踪的基本理论及关键技术研究的基础上,主要开展如下工作：
在运动目标检测方面,针对常见的基于单高斯背景模型的检测方法,提出了将判断目标点和背景点的阈值改进为自适应阈值的算法。对于基于多高斯背景模型的检测方法,提出了一种节省存储空间和降低计算复杂度的模型参数初始化方法和减少计算高斯模型的方法。最后提出了基于背景模型和帧间分析相融合的检测方法及改进。采用计算帧间图像的相关系数来代替简单的差分,根据图像中所有灰度值相同的像素点的灰度平均变化更新背景从而降低计算量,并提出了一种简单有效的消除光照条件变化方法。
在运动目标跟踪方面,对于基本的颜色分布直方图,引入HSV颜色模型代替普通的RGB颜色模型,效果更能反映人眼视觉特征。针对基本Kalman滤波的跟踪算法无法解决非线性和非高斯的问题,提出了基于扩展Kalman滤波的跟踪算法,将非线性问题线性化为的匀加速运动。
对基于Mean Shift的目标跟踪算法中的目标模型描述进行改进,不仅将基本的颜色直方图进行了核函数加权处理,还引入了边缘直方图,并对这两个特征信息进行了融合操作。对于跟踪窗口尺度不变的问题,本文也给出了解决该问题的尺度自适应算法。
针对基于粒子滤波的目标跟踪算法中粒子退化的问题,本文引入了人工免疫的思想,将目标的直方图分布作为抗原,粒子集对应的直方图分布作为抗体,目标和粒子集对应的直方图分布之间的相似性系数作为抗原和抗体的亲和力,通过克隆的方式促进亲和力大的粒子,而抑制亲和力小的粒子,使用变异的方式对亲和力小的粒子变异量大,对亲和力大的粒子变异量小,从而使系统快速收敛到全局最优解。
Motion object detection and tracking is the crucial subject of computer vision. Many fields need the technology's application, such as human-machine interface, medical diagnosis, intelligent video surveillance and intelligent robots, etc. Many subject's knowledge is needed in the research including image processing, pattern recognition and artificial intelligence.
This paper mainly does the following research in basis of the work on basic theory and key technology of motion object detection and tracking.
In the aspect of motion object detection, the paper proposes the algorithm of self-adaptive threshold which improve on the judgment threshold of object pixel and background pixel, according to the common detection method based on single Gaussian background model. The paper also proposes a method of parameter initialization that can save storage space and reduce the computational complexity, and the reduction of computation in Gaussian model. In the end of the part, a detection method and its improvement are proposed based on the fusion of background model and frame difference analysis. The simple frame difference is replaced by the correlation coefficient of the frame image, the background's update is on the basis of all the same gray value pixels' average variation to reduce the computation, and a simple and efficient method of eliminating light's change is provided.
In the aspect of motion object tracking, according to the basic color histogram, the result better reflects the human visual characteristics after introduce HSV color model instead of common RGB color model. For the problem that basic Kalman filter's tracking algorithm can't solve nonlinear and non-Gaussian operation, the tracking algorithm based on extended Kalman filter is proposed to transform nonlinear problem into the uniformly accelerated motion.
To improve the target model description of basic Mean Shift object tracking algorithm, not only the color histogram is weighted by kernel function, but also the edge histogram is introduced and the two feature information are fused. For the problem of tracking window scale invariant, the adaptive scale algorithm is proposed too.
According to the particle degradation of object tracking algorithm based on particle filter, the paper introduces the thought of artificial immune, put the target's histogram as antigen, the particle set's histogram as antibody, the similarity coefficient between the target's and particle set's histogram as the affinity between antigen and antibody, by the way of colon to promote the large affinity particle and inhibit the small affinity particle, use the mutation means to set large variation for small affinity particle and set small variation for large affinity particle, so the system can converge to the global optimal solution rapidly.
本文对运动目标的检测与跟踪的基本理论及关键技术研究的基础上,主要开展如下工作：
在运动目标检测方面,针对常见的基于单高斯背景模型的检测方法,提出了将判断目标点和背景点的阈值改进为自适应阈值的算法。对于基于多高斯背景模型的检测方法,提出了一种节省存储空间和降低计算复杂度的模型参数初始化方法和减少计算高斯模型的方法。最后提出了基于背景模型和帧间分析相融合的检测方法及改进。采用计算帧间图像的相关系数来代替简单的差分,根据图像中所有灰度值相同的像素点的灰度平均变化更新背景从而降低计算量,并提出了一种简单有效的消除光照条件变化方法。
在运动目标跟踪方面,对于基本的颜色分布直方图,引入HSV颜色模型代替普通的RGB颜色模型,效果更能反映人眼视觉特征。针对基本Kalman滤波的跟踪算法无法解决非线性和非高斯的问题,提出了基于扩展Kalman滤波的跟踪算法,将非线性问题线性化为的匀加速运动。
对基于Mean Shift的目标跟踪算法中的目标模型描述进行改进,不仅将基本的颜色直方图进行了核函数加权处理,还引入了边缘直方图,并对这两个特征信息进行了融合操作。对于跟踪窗口尺度不变的问题,本文也给出了解决该问题的尺度自适应算法。
针对基于粒子滤波的目标跟踪算法中粒子退化的问题,本文引入了人工免疫的思想,将目标的直方图分布作为抗原,粒子集对应的直方图分布作为抗体,目标和粒子集对应的直方图分布之间的相似性系数作为抗原和抗体的亲和力,通过克隆的方式促进亲和力大的粒子,而抑制亲和力小的粒子,使用变异的方式对亲和力小的粒子变异量大,对亲和力大的粒子变异量小,从而使系统快速收敛到全局最优解。
Motion object detection and tracking is the crucial subject of computer vision. Many fields need the technology's application, such as human-machine interface, medical diagnosis, intelligent video surveillance and intelligent robots, etc. Many subject's knowledge is needed in the research including image processing, pattern recognition and artificial intelligence.
This paper mainly does the following research in basis of the work on basic theory and key technology of motion object detection and tracking.
In the aspect of motion object detection, the paper proposes the algorithm of self-adaptive threshold which improve on the judgment threshold of object pixel and background pixel, according to the common detection method based on single Gaussian background model. The paper also proposes a method of parameter initialization that can save storage space and reduce the computational complexity, and the reduction of computation in Gaussian model. In the end of the part, a detection method and its improvement are proposed based on the fusion of background model and frame difference analysis. The simple frame difference is replaced by the correlation coefficient of the frame image, the background's update is on the basis of all the same gray value pixels' average variation to reduce the computation, and a simple and efficient method of eliminating light's change is provided.
In the aspect of motion object tracking, according to the basic color histogram, the result better reflects the human visual characteristics after introduce HSV color model instead of common RGB color model. For the problem that basic Kalman filter's tracking algorithm can't solve nonlinear and non-Gaussian operation, the tracking algorithm based on extended Kalman filter is proposed to transform nonlinear problem into the uniformly accelerated motion.
To improve the target model description of basic Mean Shift object tracking algorithm, not only the color histogram is weighted by kernel function, but also the edge histogram is introduced and the two feature information are fused. For the problem of tracking window scale invariant, the adaptive scale algorithm is proposed too.
According to the particle degradation of object tracking algorithm based on particle filter, the paper introduces the thought of artificial immune, put the target's histogram as antigen, the particle set's histogram as antibody, the similarity coefficient between the target's and particle set's histogram as the affinity between antigen and antibody, by the way of colon to promote the large affinity particle and inhibit the small affinity particle, use the mutation means to set large variation for small affinity particle and set small variation for large affinity particle, so the system can converge to the global optimal solution rapidly.
引文
[1]李培华,序列图像中运动目标跟踪方法,科学出版社,2010年,1-8
[2]B.K.Horn, B.Schunck. Determining optical flow, Artificial Intelligence, 1981.17(1):186-203
[3]D.Meyer, J.Denzler, and H.Niemann. Model Based Extraction of Articulated Objects in Image Sequences for Gait Analysis. Proceedings of IEEE International Conference on Image Processing, Santa Barbara,Califonia,1997.78-81
[4]王新余,张桂林,基于光流的运动目标实施检测方法研究,计算机工程与应用,2004年01期
[5]胡以静,李政访,胡跃明,基于光流的运动分析理论及应用,计算机测量与控制,2007.15(2)
[6]Ming Ye. Robust visual motion analysis:piecewise-smooth optical flow and motion-based detection and tracking[D].University of Washington,2002
[7]C. Wren, A. Azarbayejani, T. Darrell, A. Pentland, Pfinder:real-time tracking of the human body, IEEE Transactions on Pattern Analysis and Machine Intelligence 19(7)(1997)780-785
[8]D. Koller, J. Weber, T. Huang, J. Malik, G Ogasawara, B. Rao, S. Russel, Towards robust automatic traffic scene analysis in real-time, in:Proceedings of the International Conference on Pattern Recognition, Israel, November 1994.
[9]S. Jabri, Z. Duric, H. Wechsler, A. Rosenfeld, Detection and location of people in video images using adaptive fusion of colour and edge information, ICPR 2000, 3-8 September, Barcellona, Spain, pp.627-630.
[10]A. Elgammal, D. Harwood, L. Davis, Non-parametric model for background subtraction, in:IEEE Frame Rate Workshop,21 September, Corfu', Greece,1999.
[11]I. Haritaoglu, D. Harwood, L. Davis, W4:real-time surveillance of people and their activities, IEEE Transactions on Pattern Analysis and Machine Intelligence 22 (8) (2000) 809-830.
[12]K. Kim, T.H. Chalidabhongse, D. Harwood, L.S. Davis, Real-time foreground-background segmentation using codebook model, Real-Time Imaging 11 (3) (2005) 172-185.
[13]K. Toyama, J. Krumm, B. Brumitt, B. Meyers, Wallflower:principles and practice of background maintenance, in:Proceedings of the Seventh International Conference on Computer Vision,20-27 September, Corfu', Greece,1999, pp. 255-261.
[14]E. Grimson, C. Stauffer, R. Roman, L. Lee, Using adaptive tracking to classify and monitoring activities in a site, in:Computer Vision and Pattern Recognition Conference,23-25 June, Santa Barbara, CA, USA,1998, pp.22-29.
[15]C. Stauffer, W.E.L. Grimson, Learning patterns of activity using real-time tracking, IEEE Transactions on Pattern Analysis and Machine Intelligence 22 (8) (2000) 747-757.
[16]N. Friedman, S. Russel, Image segmentation in video sequences:a probabilistic approach, in:Uncertainty in Artificial Intelligence,1-3 August, Providence, RI, USA,1997.
[17]T. Kanade, T. Collins, A. Lipton, Advances in cooperative multi-sensor video surveillance Darpa Image Understanding Workshop, Morgan Kaufmann, Los Altos, CA,1998. pp.3-24.
[18]Quen-Zong Wu, Bor-Shenn Jeng, Background subtraction based on logarithimc intensities, Pattern Recognition Letters 23 (2002) 1529-1536.
[19]A. Monnet, A. Mittal, N. Paragios, V. Ramesh, Background modeling and subtraction of dynamic scenes, in:Proceedings of the International Conference on Computer Vision,2003, pp.1305-1312.
[20]A. Mittal, N. Paragios,'Motion-based background subtraction using adaptive kernel density estimation, in:Proceedings of the International Conference on Computer Vision and Pattern Recognition (CVPR),2004, pp.302-309.
[21]L. Li, W. Huang, I.Y.H. Gu, Q. Tian, Statistical modeling of complex backgrounds for foreground object detection, IEEE Transactions on Image Processing 13 (11) (2004).
[22]Q.Z. Wu, H.Y. Cheng, B.S. Jeng, Motion detection via change-point detection for cumulative histograms of ratio images, Pattern Recognition Letters 26 (5) (2005) 555-563.
[23]Y. Ivanov, A. Bobick, J. Liu, Fast lighting independent background subtraction, in: Proceedings of the IEEE Workshop on Visual Surveillance,2 January, Bombay, India,1998, pp.49-55.
[24]S.N. Lim, A. Mittal, L.S. Davis, N. Paragios, Fast illuminationinvariant background subtraction using two views:error analysis, sensor placement and applications, in:Proceedings of the International Conference on Computer Vision and Pattern Recognition (CVPR),2005, pp.1071-1078.
[25]E. Herrero-Jaraba, C. Orrite-Urunuela, J. Senar, Detected motion classification with a double-background and a neighborhood-based difference, Pattern Recognition Letters 24 (2003) 2079-2092.
[26]T. Thongkamwitoon, S. Aramvith, T.H. Chalidabhongse, Non-linear learning factor control for statistical adaptive background subtraction algorithm, Proceedings of the IEEE International Symposium on Circuits and Systems (ISCAS'05), May 23-26, Kobe, Japan,2005, pp.3785-3788
[27]Y.Kameda and M.Minoh,"A human motion estimation method using 3-successive video frames,"in Proc. Int. Conf. on Virtual Syst. And Multimedia(ICVSM),Sep.1996,pp.135-140
[28]Collins R et at. A system for video surveillance and monitoring:VSAM final report. Carnegie Mellon University. Technical Report:CMU-RI-TR-00-12,2000
[29]Shahinfard E., Sid-Ahmed M. A., Ahmadi M. A motion adaptive de-interlacing method with hierarchical motion detection algorithm. ICIP 2008.15th IEEE:889-892
[30]Crnojevic V., Antic B., Culibrk D. Optimal wavelet differencing method for robust motion detection. ICIP,2009 16th IEEE
[31]Comaniciu D., Ramesh V, Meer P. Kernel-based object tracking. IEEE Transactions on Pattern Analysis and Machine Intelligence.2003,25(5):564-577
[32]Comaniciu D., Ramesh V, Meer P. Real-time tracking of non-rigid objects using Mean Shift. Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition, New York:IEEE Press,2000,2:142-149
[33]P. Spagnolo, T.D'Orazio, M. Leo, A. Distante. Moving object segmentation by background subtraction and temporal analysis. Image and Vision Computing, 2006,24:411-423
[34]朱志宇,粒子滤波算法及其应用,科学出版社,2010年,3-8
[35]Gordon N J, Salmond D J, Smith A F M. A novel approach to nonlinear and non-Gaussian Bayesian state estimation. IEEE Proceeding,1993,140:107-113
[36]Doucet A, Godsill S, Andrieu C. On sequential Monte Carlo sampling methods for Bayesian filtering. Statistics and Computing,2000,10(3):197-208
[37]Crisan D, Doucet A. A survey of convergence results on particle filtering methods for practitioner, IEEE Trans, on Signal Processing,2002,50(2):736-746
[38]Luo C, Cai X C, Zhang J. Improved adaptive particle filter using adjusted variance and gradient data.2008 IEEE 10th Workshop Multimedia Signal Processing,2008:359-364
[39]Vaswani N. Particle filtering for larage dimensional state spaces with multimodal observation likelihoods. IEEE Trans.on Signal Processing,2008,56(2):432-440
[40]Chen Z. Bayesian filtering:From Kalman filters to particle filters, and beyond. Hamilton:McMaster University,2003
[41]Qi C, Bondon P. A new unscented particle filter. IEEE International Conference on Acoustics, Speech and Signal Processing,2008:3417-3420
[42]Kotecha J H, Djuric P M. Gaussian particle filter. IEEE Trans. on Signal Processing,2003,51(10):2593-2602
[43]Kotecha J H, Djuric P M. Gaussian sum particle filter. IEEE Trans. on Signal Processing,2003,51(10):2602-2611
[44]Kamel H, Badawy W. Fuzzy-logic-based particle filter for tracking a maneuverable target, Proc. Midwest Symp. Circuits System,2005:1537-1540
[45]Fox D. Adaptiving the sample size in particle filters through KLD-sample. Washington:University of Washington,2003:1-27
[46]Schoh T, Gustafsson F, Nordlund P J. Marginalized particle filters for mixed linear/nonlinear state-space models. IEEE Trans. on Signal Processing, 2005,53(7):2279-2289
[47]Baziw E. Real-time seismic signal enhancement utilizing a hybrid Rao-Blackwellized particle filter and hidden Markov model filter. IEEE Geoscience and Romote Sensing Letters,2005,2(4):418-422
[48]Leandro Nunes de Castro,Fernando J. Von Zuben. The clonal selection algorithm with engineering applications [C]. Workshop on Artificial Immune Systems and Their Application, Las Vegas, USA, July 2000. In Workshop Proceedings of GECCO,2000,36-37
[2]B.K.Horn, B.Schunck. Determining optical flow, Artificial Intelligence, 1981.17(1):186-203
[3]D.Meyer, J.Denzler, and H.Niemann. Model Based Extraction of Articulated Objects in Image Sequences for Gait Analysis. Proceedings of IEEE International Conference on Image Processing, Santa Barbara,Califonia,1997.78-81
[4]王新余,张桂林,基于光流的运动目标实施检测方法研究,计算机工程与应用,2004年01期
[5]胡以静,李政访,胡跃明,基于光流的运动分析理论及应用,计算机测量与控制,2007.15(2)
[6]Ming Ye. Robust visual motion analysis:piecewise-smooth optical flow and motion-based detection and tracking[D].University of Washington,2002
[7]C. Wren, A. Azarbayejani, T. Darrell, A. Pentland, Pfinder:real-time tracking of the human body, IEEE Transactions on Pattern Analysis and Machine Intelligence 19(7)(1997)780-785
[8]D. Koller, J. Weber, T. Huang, J. Malik, G Ogasawara, B. Rao, S. Russel, Towards robust automatic traffic scene analysis in real-time, in:Proceedings of the International Conference on Pattern Recognition, Israel, November 1994.
[9]S. Jabri, Z. Duric, H. Wechsler, A. Rosenfeld, Detection and location of people in video images using adaptive fusion of colour and edge information, ICPR 2000, 3-8 September, Barcellona, Spain, pp.627-630.
[10]A. Elgammal, D. Harwood, L. Davis, Non-parametric model for background subtraction, in:IEEE Frame Rate Workshop,21 September, Corfu', Greece,1999.
[11]I. Haritaoglu, D. Harwood, L. Davis, W4:real-time surveillance of people and their activities, IEEE Transactions on Pattern Analysis and Machine Intelligence 22 (8) (2000) 809-830.
[12]K. Kim, T.H. Chalidabhongse, D. Harwood, L.S. Davis, Real-time foreground-background segmentation using codebook model, Real-Time Imaging 11 (3) (2005) 172-185.
[13]K. Toyama, J. Krumm, B. Brumitt, B. Meyers, Wallflower:principles and practice of background maintenance, in:Proceedings of the Seventh International Conference on Computer Vision,20-27 September, Corfu', Greece,1999, pp. 255-261.
[14]E. Grimson, C. Stauffer, R. Roman, L. Lee, Using adaptive tracking to classify and monitoring activities in a site, in:Computer Vision and Pattern Recognition Conference,23-25 June, Santa Barbara, CA, USA,1998, pp.22-29.
[15]C. Stauffer, W.E.L. Grimson, Learning patterns of activity using real-time tracking, IEEE Transactions on Pattern Analysis and Machine Intelligence 22 (8) (2000) 747-757.
[16]N. Friedman, S. Russel, Image segmentation in video sequences:a probabilistic approach, in:Uncertainty in Artificial Intelligence,1-3 August, Providence, RI, USA,1997.
[17]T. Kanade, T. Collins, A. Lipton, Advances in cooperative multi-sensor video surveillance Darpa Image Understanding Workshop, Morgan Kaufmann, Los Altos, CA,1998. pp.3-24.
[18]Quen-Zong Wu, Bor-Shenn Jeng, Background subtraction based on logarithimc intensities, Pattern Recognition Letters 23 (2002) 1529-1536.
[19]A. Monnet, A. Mittal, N. Paragios, V. Ramesh, Background modeling and subtraction of dynamic scenes, in:Proceedings of the International Conference on Computer Vision,2003, pp.1305-1312.
[20]A. Mittal, N. Paragios,'Motion-based background subtraction using adaptive kernel density estimation, in:Proceedings of the International Conference on Computer Vision and Pattern Recognition (CVPR),2004, pp.302-309.
[21]L. Li, W. Huang, I.Y.H. Gu, Q. Tian, Statistical modeling of complex backgrounds for foreground object detection, IEEE Transactions on Image Processing 13 (11) (2004).
[22]Q.Z. Wu, H.Y. Cheng, B.S. Jeng, Motion detection via change-point detection for cumulative histograms of ratio images, Pattern Recognition Letters 26 (5) (2005) 555-563.
[23]Y. Ivanov, A. Bobick, J. Liu, Fast lighting independent background subtraction, in: Proceedings of the IEEE Workshop on Visual Surveillance,2 January, Bombay, India,1998, pp.49-55.
[24]S.N. Lim, A. Mittal, L.S. Davis, N. Paragios, Fast illuminationinvariant background subtraction using two views:error analysis, sensor placement and applications, in:Proceedings of the International Conference on Computer Vision and Pattern Recognition (CVPR),2005, pp.1071-1078.
[25]E. Herrero-Jaraba, C. Orrite-Urunuela, J. Senar, Detected motion classification with a double-background and a neighborhood-based difference, Pattern Recognition Letters 24 (2003) 2079-2092.
[26]T. Thongkamwitoon, S. Aramvith, T.H. Chalidabhongse, Non-linear learning factor control for statistical adaptive background subtraction algorithm, Proceedings of the IEEE International Symposium on Circuits and Systems (ISCAS'05), May 23-26, Kobe, Japan,2005, pp.3785-3788
[27]Y.Kameda and M.Minoh,"A human motion estimation method using 3-successive video frames,"in Proc. Int. Conf. on Virtual Syst. And Multimedia(ICVSM),Sep.1996,pp.135-140
[28]Collins R et at. A system for video surveillance and monitoring:VSAM final report. Carnegie Mellon University. Technical Report:CMU-RI-TR-00-12,2000
[29]Shahinfard E., Sid-Ahmed M. A., Ahmadi M. A motion adaptive de-interlacing method with hierarchical motion detection algorithm. ICIP 2008.15th IEEE:889-892
[30]Crnojevic V., Antic B., Culibrk D. Optimal wavelet differencing method for robust motion detection. ICIP,2009 16th IEEE
[31]Comaniciu D., Ramesh V, Meer P. Kernel-based object tracking. IEEE Transactions on Pattern Analysis and Machine Intelligence.2003,25(5):564-577
[32]Comaniciu D., Ramesh V, Meer P. Real-time tracking of non-rigid objects using Mean Shift. Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition, New York:IEEE Press,2000,2:142-149
[33]P. Spagnolo, T.D'Orazio, M. Leo, A. Distante. Moving object segmentation by background subtraction and temporal analysis. Image and Vision Computing, 2006,24:411-423
[34]朱志宇,粒子滤波算法及其应用,科学出版社,2010年,3-8
[35]Gordon N J, Salmond D J, Smith A F M. A novel approach to nonlinear and non-Gaussian Bayesian state estimation. IEEE Proceeding,1993,140:107-113
[36]Doucet A, Godsill S, Andrieu C. On sequential Monte Carlo sampling methods for Bayesian filtering. Statistics and Computing,2000,10(3):197-208
[37]Crisan D, Doucet A. A survey of convergence results on particle filtering methods for practitioner, IEEE Trans, on Signal Processing,2002,50(2):736-746
[38]Luo C, Cai X C, Zhang J. Improved adaptive particle filter using adjusted variance and gradient data.2008 IEEE 10th Workshop Multimedia Signal Processing,2008:359-364
[39]Vaswani N. Particle filtering for larage dimensional state spaces with multimodal observation likelihoods. IEEE Trans.on Signal Processing,2008,56(2):432-440
[40]Chen Z. Bayesian filtering:From Kalman filters to particle filters, and beyond. Hamilton:McMaster University,2003
[41]Qi C, Bondon P. A new unscented particle filter. IEEE International Conference on Acoustics, Speech and Signal Processing,2008:3417-3420
[42]Kotecha J H, Djuric P M. Gaussian particle filter. IEEE Trans. on Signal Processing,2003,51(10):2593-2602
[43]Kotecha J H, Djuric P M. Gaussian sum particle filter. IEEE Trans. on Signal Processing,2003,51(10):2602-2611
[44]Kamel H, Badawy W. Fuzzy-logic-based particle filter for tracking a maneuverable target, Proc. Midwest Symp. Circuits System,2005:1537-1540
[45]Fox D. Adaptiving the sample size in particle filters through KLD-sample. Washington:University of Washington,2003:1-27
[46]Schoh T, Gustafsson F, Nordlund P J. Marginalized particle filters for mixed linear/nonlinear state-space models. IEEE Trans. on Signal Processing, 2005,53(7):2279-2289
[47]Baziw E. Real-time seismic signal enhancement utilizing a hybrid Rao-Blackwellized particle filter and hidden Markov model filter. IEEE Geoscience and Romote Sensing Letters,2005,2(4):418-422
[48]Leandro Nunes de Castro,Fernando J. Von Zuben. The clonal selection algorithm with engineering applications [C]. Workshop on Artificial Immune Systems and Their Application, Las Vegas, USA, July 2000. In Workshop Proceedings of GECCO,2000,36-37