( 您好!臺灣時間:2021/07/29 14:05
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::


論文名稱(外文):Study on Moving Target Detection and Tracking Based on Particle Filter
指導教授(外文):Ming-Yang Cheng
外文關鍵詞:vision surveillance systemparticle filteractive visual tracking
  • 被引用被引用:0
  • 點閱點閱:382
  • 評分評分:
  • 下載下載:37
  • 收藏至我的研究室書目清單書目收藏:0
基於視覺之智慧型安全監控系統近年來廣受矚目,其研究範圍包含電腦視覺演算法與攝影機之間的溝通等等。本論文中以一PTU攝影機結合個人電腦實現一主動式視覺監控系統,此系統分為移動物偵測與目標物追蹤兩個部份。在移動物偵測方面,使用改良式適應性背景減去法,能夠克服背景模型變化及物體移動緩慢或暫時靜止的情況。在追蹤方面,以顏色直方圖建立目標物模型,並使用近年來在各領域廣為應用的粒子濾波法,並將之應用於主動式追蹤。此外,在本論文中亦介紹另外兩種改良式的粒子濾波法,分別是核心粒子濾波法(KPF)以及卡爾曼粒子濾波法(Kalman PF)。在KPF演算法中以核心密度估測事後機率密度函數,並利用平均位移法將粒子沿著核心密度函數的梯度方向移動,屬於一種爬峰式的架構;另一種則是在粒子濾波法中加入卡爾曼濾波法,在最終狀態估測之前,使用卡爾曼濾波法中的更新方程式,使粒子移動至概似函數較高的位置。實驗中以一個放置於類似倒單擺機構的圓球模擬物體進行不規則運動,利用一PTU攝影機對其進行主動式追蹤,將目標鎖定在畫面中央,並比較三種演算法的追蹤表現,且使用不同粒子數進行實驗,比較三種演算法性能上的變化。
In recent years, vision surveillance and monitoring systems have attracted much attention, in which the related research topics include computer vision and communication among cameras. In this thesis, a PTU camera combined with a personal computer is used to implement an active visual surveillance system that consists of two parts  moving object detection and target tracking. Modified adaptive background subtraction which can deal with the problems such as the change in background model and the situation that the target moves slowly or temporally stops is employed in the detection part. As for target tracking, color histogram is used to construct the target model, and the particle filter which is very popular in various research fields is applied in active tracking. Besides, two types of modified particle filters  Kernel Particle Filter (KPF) and Kalman Particle Filter (Kalman PF) are introduced in this thesis. KPF uses Kernel density estimate (KDE) to approximate the posterior probability density function and particles move along the gradient direction of the kennel density function using the mean shift method, which is regarded as a mode-seeking approach. Kalman PF is a particle filter that exploits the principle of Kalman filter, which moves particles to the position with higher likelihood using the update equation of the Kalman Filter. In order to compare the performance of three different tracking algorithms, several tracking experiments are conducted. In the experiment, a ball attached to an inverted pendulum-like mechanism is used to simulate the scenario that an object undergoes an irregular motion, while the active visual surveillance system consisting of a PTU camera and a PC that executes different tracking algorithms is used to track the motion of the ball so as to lock its image at the center of the monitoring screen. In addition, the numbers of particles are varied in different tracking experiments to compare the performance of different algorithms.
中文摘要 III
誌謝 VI
目錄 VII
表目錄 X
圖目錄 XI
第一章 緒論 1
1.1 研究動機與目的 1
1.2 文獻回顧 4
1.3 本文架構 5
第二章 移動目標物偵測 6
2.1 前景偵測 6
2.2 適應性背景減去法 9
2.3 改良式適應性背景減去法 11
2.3.1 物件標示 12
2.3.2 重疊分類法 12
2.3.3 前景相似度 13
2.3.4 靜止物體之背景更新 18
第三章 粒子濾波法 20
3.1 貝氏濾波法 20
3.2 卡爾曼濾波法 23
3.3 粒子濾波法 25
3.3.1 SIS演算法 25
3.3.2 SIR演算法 27
第四章 以粒子濾波器為基礎之目標物追蹤 32
4.1 以顏色為基礎之粒子濾波法 33
4.1.1 顏色之量測 33
4.1.2 演算法流程 37
4.1.3 目標物模型更新 40
4.2 以顏色為基礎之卡爾曼粒子濾波法 40
4.2.1 系統模型與概似函數 41
4.2.2 粒子狀態與觀測模型 43
4.2.3 Kalman PF演算法流程 45
4.3 結合平均位移法之粒子濾波法 48
第五章 實驗結果與討論 54
5.1 軟硬體設備概述 54
5.2 實驗結果與討論 61
5.2.1 實驗一:粒子數50 65
5.2.2 實驗二:粒子數20 74
5.2.3 綜合分析 82
第六章 結論與建議 89
6.1 結論 89
6.2 未來展望與建議 91
參考文獻 92
[1]B. K. P. Horn and B. G. Schunck, Determining optical flow, Artificial intelligence, vol. 17, pp. 185-203, 1981.
[2]B. D. Lucas and T. Kanade, An iterative image registration technique with an application to stereo vision, in Proceedinds of the International Joint Conferenrce on Artificial Intelligence, pp .674-679, 1981.
[3]R.T. Collins, A. J. Lipton, T. Kanade, H. Fujiyoshi, D. Duggins, Y. Tsin, D. Tolliver, N. Enomoto, O. Hasegawa, P. Burt and L. Wixson, “A system for surveillance and monitoring, The Robotics Institute, Carnegie Mellon University, Pittsburgh PA, Tech. Rep. CMU-RI-TR-00-12, 2000.
[4]王俊凱, 基於改良式適應性背景相減法與多重影像特徵比對法之多功能即時視覺追蹤系統之設計與實現, 碩士論文, 電機工程學系, 國立成功大學, 2004.
[5]E. Davies, Machine Vision: Theory, Algorithms and Practicalities, Academic Press, 1990.
[6]M. Sonka, V. Hlavac and R. Boyle, Image Processing, Analysis, and Machine Vision. Pacific Grove: PWS, 1999
[7]J. Hill and W. T. Park, Real time control of a robot with a mobile camera, in Proceedings of 9th ISIR, pp. 233-246, 1979.
[8]R. C. Gonzalez and R. E. Woods, “Digital Image Processing, New Jersey: Prentice-Hall, 2002.
[9]M. Sanjeev Arulampalam, Simon Maskell, Neil Gordon, and Tim Clapp, “A tutorial on Particle Filters for Online Nonlinear/Non-Gaussian Bayesian Tracking, IEEE Transactions on Signal Processing, vol.50,pp.174-188, 2002.
[10]A. Doucet, J. F. G. de Freitas, and N. J. Gordon, “An introduction to sequential Monte Carlo methods, in Sequential Monte Carlo Methods in Practice, A. Doucet, J. F. G. de Freitas, and N. J. Gordon, Eds. New York: Springer-Verlag, 2001.
[11]A. Doucet, S. Godsill, and C. Andrieu, “On sequential Monte Carlo sampling methods for Bayesian filtering, Statist.Comput., vol. 10, no. 3, pp.197–208.
[12]N. Gordon, D. Salmond, and A. F. M. Smith, “Novel approach to nonlinear and non-Gaussian Bayesian state estimation, Proc. Inst. Elect.Eng., F, vol. 140, pp. 107–113, 1993.
[13]Michael Isard and Andrew Blake, “CONDENSATION-Conditional Density Propagation for Visual Tracking,International Journal of Computer Vision,vol.29,
[14]N. Gordon, D. Salmond, and A. F. M. Smith, “Novel approach to nonlinear and non-Gaussian Bayesian state estimation, Proc. Inst. Elect.Eng., F, vol. 140, pp. 107–113, 1993.
[15]J. Carpenter, P. Clifford, and P. Fearnhead, “Improved particle filter for nonlinear problems, Proc. Inst. Elect. Eng., Radar, Sonar, Navig.,pp.2-7, 1999.
[16]A. Doucet, “On sequential Monte Carlo methods for Bayesian filtering,Dept. Eng., Univ. Cambridge, UK, Tech. Rep., 1998.
[17]S. Thrun, W. Burgard, and D. Fox, Probability Robotics, The MIT Press, 2005.
[18]G. Kitagawa, “Monte Carlo filter and smoother for non-Gaussian nonlinear state space models, J. Comput. Graph. Statist., vol. 5, no. 1, pp.1–25, 1996.
[19]R. E. Kalman and R. S. Bucy, New Results in Linear Filtering and Prediction Theory, Transactions of the ASME – Journal of Basic Engineering, vol. 83, pp. 95-107, 1961.
[20]Y. C. Ho and R. C. K. Lee, “A Bayesian approach to problems in stochastic estimation and control, IEEE Trans. Automat. Contr., pp. 333–339, 1964.
[21]K. Nummiaro, E. Koller-Meier, and L. Van Gool, An Adaptive Color-Based Particle Filter, International Journal of Image and Vision Computing, vol. 21, pp. 99-110, 2003.
[22]P. Li et al., “Visual contour tracking based on particle filters, Image Vision Computing, vol.21, pp.111–123, 2003.
[23]Yoshinori Satoh, Takayuki Okatani and Koichiro Deguchi,A color-based Tracking by Kalman Particle Filter, Pattern Recognition, 2004. ICPR 2004.
[24]Cheng Chang, Rashid Ansari,Kernel Particle Filter for Visual Tracking, IEEE Signal Processing Letters, vol.12,pp242-245,2005
[25]D. Comaniciu, V. Ramesh, and P. Meer, “Kernel-based object tracking, IEEE Trans. Pattern Anal. Mach. Intell., vol. 25, no. 5, pp. 564–577, May 2003.
[26]D. Comaniciu and P. Meer, “Mean Shift: A Robust Approach Toward Feature Space Analysis, IEEE Trans. Pattern Analysis and Machine Intelligence, vol.24, no.5, pp.603-619,May 2002.
[27]J. M. Hammersley and K. W. Morton, “Poor man’s Monte Carlo, Journal of the Royal Statistical Society, vol. 16, pp.23-28, 1954.
[28]M. West and J. Harrison, “Bayesian forecasting and dynamic models, Springer Series in Statistics, 2nd ed. New York: Springer-Verlag, 1997.
[29]J.Carpenter, P.Clifford, P.Fearnhead, “Improved Particle filter for nonlinearproblems, IEE Proc. Radar Sonar Navig., vol. 146, pp.2-7.
[30]N.Gordon, D.Salmond, A.F.M.Smith, “Novel approach to nonlinear and non-Gaussian Bayesian state estimation, IEE Proceedings-F , vol. 140, pp.107-113, 1993.
[31]K.Kanazawa, D.Koller, S.J.Russell, “Stochastic simulation algorithms for dynamic probabilistic networks, Proc. 11th Annul Conf. on Uncertainty AI, pp. 366-351, 1995.
[32]J.MacCormick, A.Blake, A probabilistic exclusion principle for tracking multiple objects. In Proc. 7th IEEE Internat. Conf. on Computer Vision, vol. 1, pp. 572-578, 1999.
[33]D.Comaniciu, V.Ramesh, ,Mean shift and optimal prediction for efficient object tracking, IEEE Internat. Conf. on Image Processing, Vancouver, Canada, pp. 70-73, 2000.
[34]D.Comaniciu, V.Ramesh, P.Meer, “Real-time tracking of non-rigid objects using mean shift, IEEE Conf. on Computer Vision and Pattern Recognition,Hilton Head, SC, pp.142-149, 2000.
[35]C.Shan, T.Tan, Y.Wei, “Real time hand tracking using a mean shift embedded particle filter,Pattern Recognition, vol. 40, pp.1958-1970, 2007.
[36]C.Chang and R.Ansari, “Kernel Particle Filter for Visual Tracking,IEEE Signal Processing Letters, vol. 12 ,pp.242-245, 2005.
[37]Parisa Darvish Zadeh Varcheie and Guillaume-Alexandre Bilodeau, “Adaptive Fuzzy Particle Filter Tracker for a PTZ Camera in an IP Surveillance System, IEEE Transactions on Instrumentation and Measurement, vol. 60, pp.354-371, 2011.
[38]P. Pan and D. Schonfeld, “Visual Tracking Using High-Order Particle Filtering, IEEE Signal Processing Letters, vol.18, pp.51-54, 2011.
[39]Y.Zhai, Mark B. Yeary, S.Cheng and N.Kehtarnavaz, “An Object-Tracking Based on Multi-Model Particle Filtering With State Partitioning, IEEE Transactions on Instrumentation and Measurement, vol.58, pp.1797-1809, 2009.
[40]T.Balch, F.Dellaert, “MCMC-based particle filtering for tracking a variable number of interacting targets, Pattern Analysis and Machine Intelligence, vol.27, pp.1805-1819, 2005.
[41]P. Fearnhead, “Particle filters for mixture models with an unknown number of components, Statistics and Computing, vol.14, pp.11-21, 2004.
[42]Peter I.Corke and Malclom C.Good, “Dynamic Effects in Visual Closed-Loop Systems, IEEE Transactions on Robotics and Automation, vol 12, 1996.

註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
第一頁 上一頁 下一頁 最後一頁 top