臺灣博碩士論文加值系統

影像追蹤的應用向來都是一個相當重要的議題，早期曾運用在空中交通控制，近年來則多有應用在安全監控管理方面的需求。目標追蹤的方法有許多類型，大致可分為在時域、或空間域上的操作。在時域操作的系統中，物件必須要在時間上反映出差異，也就是物件要有移動才有辦法做判斷。而空間域操作的系統則是由影像中物件的特徵資訊來做判斷，通常特徵資訊判斷的方法較複雜且多元。然而，單純的使用時域的方法只能偵測到有物件在移動，並不能知道該物件是否為使用者有興趣的物件。場景必須為已知並且要預先設定物件物可能出現的位置，方能適用，然而，藉由Adaboost分類演算法的加入即可解決此一問題。
有鑒於此，本文提出了結合使用Adaboost架構與粒子濾波器的方式來進行自動化追蹤物件的方法，以彌補現行使用粒子濾波器進行追蹤的方法中，需要預先於場景中指定目標，方能完成追蹤的缺陷。本文先以Adaboost進行可疑物件偵測、過濾並定位後，再以粒子濾波器進行目標物確認並追蹤，並交替使用Adaboost檢測以對粒子濾波器的追蹤結果進行校正，以防止物件的遺失。經由實驗結果顯示，與現行的追蹤法相較，本論文提出的架構在物件消失、遮蔽、再出現等情形下，皆能更成功地達成物件追蹤的目標。
然而，本論文所提出的目標物追蹤法，仍與現行的其他方法一樣，存有在複雜背景環境下，偵測及追蹤不佳的問題。在未來工作中，我們將藉由新增物件偵測前的影像強化前處理步驟，以及增加cascade訓練的負樣本數目以濾除錯誤的物件樣本檢出數，來解決上述問題，使得本系統的應用得以更廣，且更為有效。

Application of object tracking has always been an important issue in computer vision or image processing applications. In the early stages, object tracking had been applied to air traffic control. Recently, it has often been applied to with security monitoring related fields. There are various types of methods for object tracking. Generally, these methods can be divided into time domain methods and space domain methods. In a time domain system, the target must be able to show time differences. In other words, the target has to move so that judgments can be made. In a space domain system, judgments are made based on image characteristics of the targets. And usually judging methods based on characteristic information are more complex and diversified. However, if only a time domain method is applied, the only thing that can be confirmed is that whether the target is moving or not. It is difficult to find out if this target is the interest one. On the other hand, particle filter is a good solution for target tracking but is only applicable when the scene is known and possible locations of the target are preset. However, adding the adaboost algorithm helps to solve this issue. Therefore, this study proposed a combined structure with adaboost detection and particle filtering method to resolve the problems mentioned above for the pedestrian tracking problems.
Considering this problem, a hybrid structure combining adaboost classifier and particle filter is proposed to automatically detect and track the pedestrian targets in this paper. The adaboost detection process is adopted first to target candidate objects, and then the particle filter is applied for confirming and tracking of targets. Experiment results show that via the proposed method, the drawback of the current particle filters which requires specifying an object to be tracked in advance can be overcome, while performing good also in cases of target missing, occlusion, and identifying the previously appeared objects.
Like other current methods, the issue of bad performances in detection and tracking in a complex environment still exists with the object tracking method proposed by this study. In the future, we will add a preprocessing step of image enhancement before target detection and increase the number of negative samples in the sample for cascade training, to solve the issue above, so this system can be applied more widely with efficiency.

第一章 緒論 1
1.1 研究動機與背景 1
1.2 相關研究 2
1.3 論文結構 4
第二章 AdaBoost介紹 5
2.1 AdaBoost演算法簡介 5
2.2 訓練流程 7
2.2.1 Haar-like矩形特徵 7
2.2.2 積分圖(Integral Image) 10
2.2.3 任意矩形區域內像素積分 12
2.2.4 計算特徵值 13
2.2.5 抽取Haar特徵 16
2. 2. 6 產生弱分類器 17
2. 2. 7 用AdaBoost演算法選取最佳化的弱分類器 18
2. 2. 8 組成串聯結構的層疊分類器 20
2.3 AdaBoost偵測流程 22
第三章 粒子濾波器介紹 23
3.1 粒子濾波器簡介 23
3.2 動態空間模型 24
3.3 卡爾曼(Kalman)濾波器 25
3.4 Bayes理論和估計 28
3.5 蒙特卡羅(Monte Carlo)積分 31
3.6 連續蒙特卡羅(Monte Carlo)訊號處理 33
3.7 重新取樣技術 36
3.8 粒子濾波器 40
第四章 系統架構 46
4.1 Adaboost物件偵測 47
4.2 Motion Estimation 50
4.3 Adaboost & Motion Estimation 目標定位 51
4.4 粒子濾波器追蹤 52
4.5 Adaboost & 粒子濾波器交互參照校正 56
4.6 流程圖 57
第五章 實驗結果與討論 60
5.1 實驗結果 60
5.2 討論與建議 78
第六章 結論與未來工作 80
6.1 結論 80
6.2 未來工作 81
參考文獻 83
發表論文 89

[1] Bilik, and J. Tabrikian, “Target tracking in glint noise environment using nonlinear non-Gaussian Kalman filter”, in Proc. of IEEE Int. Conf. Radar, pp.282-287 , Beer-Sheva 84105, Israel, Apr. 2006.
[2] Amer, “Voting-based simultaneous tracking of multiple video objects”, IEEE Trans. on Circuits and Schemes for Video Technology, vol. 15, no. 11, pp.1448-1462, Nov. 2005.
[3] E. Loutas, K. Diamantaras, and I. Pitas, “Occlusion resistant object tracking”, in Proc. of IEEE Int. Conf. on Image Processing, pp.65-68, Oct. 2003.
[4] J. Li, and C. S. Chua, “Transductive inference for color-based particle filter tracking”, in Proc. of IEEE Int. Conf. on Image Processing, pp.14-17, Oct. 2003.
[5] Z. Q. Wen, and Z. X. Cai, “Mean shift algorithm and its application in tracking of objects”, in Int. Conf. on Machine Learning and Cybernetics, pp.4024-4028, Aug. 2006.
[6] Papageorgiou, and T. Poggio, “A trainable scheme for object detection”, International Journal of Computer Vision, vol. 38, no. 1, pp.15-33, 2005.
[7] M. Gavrila, and V. Philomin, “Real-time object detection for ”smart” vehicles”, in Proc. of IEEE Int. Conf. on Computer Vision, pp.20-25, June 1999.
[8] P. Rosin, “Thresholding for change detection”, in Proc. of IEEE Int. Conf. on Computer Vision, pp.274-279, Jan. 1998.
[9] N. Friedman, and S. Russell, “Image segmentation in video sequences: a probabilistic approach”, in Proc. of the 13th conf. on Uncertainty in Articial intelligence, pp.175-181, Aug. 1997.
[10] R. Wren, A. Azarbayejani, T. Darrell, and A. Pentland, “Pfinder: real-time tracking of the human body”, IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 19, no. 7, pp.780-785, July 1997.
[11] Javed, K. Shafique, and M. Shan, “A hierarchical approach to robust background subtraction using color and gradient information”, in Proc. of IEEE Workshop on Motion and Video Computing, pp.22-27, Dec. 2002.
[12] M. Cristani, M. Bicego, and V. Murino, “Integrated region and pixel-based approach to background modeling”, in Proc. of IEEE Workshop on Motion and Video Computing, pp.3-8, Dec. 2002.
[13] C. Stauffer, and E. Grimson, “Adaptive background mixture models for real-time tracking”, in Proc. IEEE Int. Conf. on Computer Vision and Pattern Recognition, pp.246-252, 1999.
[14] S. Lee, J. J. Hull, and B. Erol, “A bayesian framework for Gaussian mixture background modeling”, in Proc. IEEE Int. Conf. on Image Processing, pp.973-976, Sept. 2003.
[15] Elgammal, D. Harwood, and L. Davis, “Non-parametric model for background subtraction”, in Proc. IEEE Int. Conf. on Computer Vision, pp.246-252, June 1999.
[16] M. Heikkila, and M. Pietikainen, “A texture-based method for modeling the background and detecting moving objects”, IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 28, no. 4, pp.657-652, Apr. 2006.
[17] L.G. Valiant, “A Theory of the learnable”, Communication of ACM, vol. 27, pp.1134–1142, 1984.
[18] Michael Kearns, “Thoughts on Hypothesis Boosting”, Unpublished manuscript, 1988.
[19] Robert E. Schapire, “The Strength of Weak Learnability”, Machine Learning, vol. 5, no. 2, pp.197-227, 1990.
[20] Yoav Freund, “Boosting a Weak Learning Algorithm by Majority”, Information and Computation, vol. 121, no. 2, pp.256-285, Sep. 1995.
[21] Yoav Freund, and Robert E. Schapire, “Experiments with a New Boosting Algorithm”, In Proceedings of the 13th International Conference on Machine Learning, Bari, Italy (ICML). pp.148-156, 1996.
[22] Viola P., and Jones M., “Robust Real-time Object Detection”, International Journal of Computer Vision, IJCV 2004, vol. 57, no. 2, pp.137-154, 2004.
[23] Rainer Lienhart, and Jochen Maydt, “An Extended Set of Haar-like Features for Rapid Object Detection”, IEEE ICIP 2002, vol. 1, pp.900-903, Sep. 2002.
[24] Rainer Lienhart, Alexander Kuranov, and Vadim Pisarevsky, “Empirical Analysis of Detection Cascades of Boosted Classifiers for Rapid Object Detection”, MRL Technical Report, Intel Labs, May 2002, revised Dec. 2002.
[25] Chang Huang, Bo Wu, Haizhou Ai, and Shihong Lao, “Omni-Directional Face Detection based on Real AdaBoost”, International Conference on Image Processing ( ICIP’04), 2004.
[26] 陳谷瑋, “多類別AdaBoost穩健性研究”, 國立中正大學碩士論文,2005.
[27] 吳旻峰, “基於像素導向之階層式特徵與統計式遮罩AdaBoost人臉偵測”, 國立臺灣科技大學碩士論文, 2008.
[28] 陳致豪, “以膚色偵測加速AdaBoost人臉偵測”, 明志科技大學 ,2010.
[29] Hammersley J. M., and Morton K. W., “Poor man’s Monte Carlo”, Jurnal of the Royal statistical (society B) (Methodological), vol. 16, pp.23-38, 1954.
[30] Handschin J. E., “Monte Carlo techniques to estimate the conditional expectation in multi-stage non-linear filtering”, International Journal of Control, vol. 9, pp.547-559, 1969.
[31] Handschin J. E., “Monte Carlo techniques for prediction and filtering of non-linear stochastic processes”, Automatica, vol. 6, pp.555-563, 1970.
[32] Gordon N. J., Salmond D. J., and Smith A. F. M., ”A novel approach to nonlinear and non-Gaussian Bayesian state estimation”, IEEE Proceeding, vol. 140, pp.107-113, 1993.
[33] Doucet A., Godsill S., and Andrieu C., “On sequential Monte Carlo sampling methods for Bayesian filtering”, Statistics and Computing, vol. 10, no. 3, pp.197-208, 2000.
[34] Kay S. M., “統計信號處理基礎―估計與檢測理論.” 羅鵬飛, 等譯. 北京:電子工業出版社, 2003.
[35] Shneider Y. A., “Method of Statistical Testing(Monte Carlo Method)”, Oxford: Pergamon Press, p.84, 1964.
[36] Wang X., Chen R., and Liu J. S., “Monte Carlo signal processing for wireless communications”, J. VLSI Signal Process, vol. 30, pp.89-105, 2002.
[37] Kong A., Liu J. S., and Wong W. H., “Sequential imputations and Bayesian missing data problems”, J. Am. Stat. Assoc., vol. 89, pp.278-288, 1994.
[38] Sanjeev M., and SimonM., “A tutorial on particle filter for online nonlinear/non-Gaussian Bayesian tracking”, IEEE trans. on Signal Processing, vol. 50, no. 2, pp.174-188, 2002.
[39] Polson N. G., Strond J. R., and Muller P., “Practical filtering with sequential parameter learning”, University of Pennsylvania Working Paper, 2006.
[40] Chyun-Chau Fuh, “Detecting unstable periodic orbits embedded in chaotic systems using the simplex method”, Communications in Nonlinear Science and Numerical simulation, no. 14, pp.1032-1037, 2009.
[41] Liu J S., and Chen R., ”Sequential Monte Carlo methods for dynamic systems”, Journal of the American Statistical Association, vol. 93, no. 443, pp.1032-1044, 1998.
[42] Carpenter J., Clifford P., and Fearnhead P., “An improved particle filter for nonlinear problems”, IEE Proc., Radar sonar Navigation. 1999, vol. 146, pp.2-7, 1999.
[43] Miodrag Bolic, Peter M Djuric, and Sangjin Hong, “New Resampling Algorithms for particle filters ”, Proc. of IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 2, pp.589-592, 2003.
[44] Doucet A, de Freitas J, and Gordon N., “Sequential Monte Carlo Methods in Parctice”, New York: Springer, pp.159-175, 2001.
[45] Doucet A, de Freitas J, and Gordon N., “An introduction to sequential Monte Carlo methods”, New York: Springer-Verlag, pp.3-14, 2001.
[46] Hu X L, Schon T B, and Ljung L., “A basic convergence result for particle filtering”, IEEE Transactions Signal Processing, vol. 56, no. 4, pp.1337-1348, 2007.
[47] Crisan D, and Doucet A. “Convergence of sequential Monte Carlo methods”, UnivCambridge, UK, Signal Process Group, Dept Eng, 2000.
[48] 朱志宇. “粒子濾波算法及其應用.” 北京：科學出版社, 2010年6月.
[49] 胡士強, 敬忠良. “粒子濾波原理及其應用.” 北京：科學出版社, 2010年8月.
[50] Nummiaro K., Koller-Meier E., and Van Gool L., “An adaptive color-based particle filter”, Image and Vision Computing, vol. 21, pp.111-123, 2003.