高斯混合模型(Gaussian Mixture Model)常被用在各種智慧型監控系統中來分離出前景與背景。此方法雖然能有效偵測出移動中之前景，但是，若移動目標物停留在畫面上過久則會被吸收成為背景的一部分，此問題會影響前景偵測的準確性，同時後續追蹤的正確性也會受影響。
本論文提出一結合卡爾曼濾波器(Kalman filter)和高斯混合模型(簡稱為KGMM)來分離前景與背景，並用它來偵測及追踪人員於警戒區域周圍的安全。本方法可有效地分離前景與背景，改善傳統高斯混合型中前景物件停留過久會融於背景的問題，用以提高前景偵測之準確性。在實驗中我們將提出的方法結合追蹤與路徑分析的技術並應用於焚化爐周圍環境的安全管理。當人員進入警戒區域或不慎掉落到危險區域時，能及時提出警報且不受到垃圾車移動或是掉落的垃圾而影響判斷。此外，亦可應用於火車月台及圍牆周圍安全等監控偵測。透過不同場景的實驗，證明本方法可以有效克服前景物件停留原地而被吸收成為背景的問題。

Gaussian mixture background model is one of the most popular motion detection methods in intelligent surveillance system for separating the foreground from the background. Although this method can detect the moving object effectively. But, the object will be considered as a component of the background, if the object stays on the same place last for a while. This problem will affect the accuracy of the following detection and tracking steps.
In this thesis, we proposed a method (named as KGMM) based on Gaussian Mixture Model with Kalman filter to solve the above problem and used it to detect and track the people around the warning area. This motion detection method can correctly detect the stopped objects and moving objects on the screen. In our experiment, we integrate it with tracking and path analysis technologies into the surveillance to monitor the safety of the surrounding areas of the incinerator. It will alarm in time when someone gets into the warning area, or falls into the dangerous area. Also, this can be used to monitor the safety of the railway platform and enclosing wall. Based on the experiments with difference scenes, the proposed method can effectively solve the foreground melted into background problem.

[1] G. Welch and G. Bishop, ”An introduction to the Kalman filter,” Working Paper, Department of Computer Science, University of North Carolina at Chapel Hill, 2006.
[2] C. Stauffer and W. E. L Grimson, “Adaptive background mixture models for real-time tracking,” in Proc. of IEEE Conference on Computer Vision and Pattern Recognition, vol. 2, pp. 246-252, 1999.
[3] J. Hill and W. T. Park, “Real time control of a robot with a mobile camera,” in Proc. 9th ISIR, pp. 233-246, 1979.
[4] 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 video surveillance and monitoring”, Tech. Rep., The Robotics Institute, Carnegie Mellon University, 2000.
[5] X. Clady, F. Collange, F. Jurie, and P. Martinet, “Cars detection and tracking with a vision sensor,” in Proc. of the IEEE on Intelligent Vehicles Symposium, pp. 593 – 598, June. 2003.
[6] Van Leeuwen, M.B. and Groen, F.C.A., "Vehicle detection with a mobile camera: spotting midrange, distant, and passing cars,” Robotics & Automation Magazine, IEEE, Vol. 12 Issue 1, pp. 37-43, March. 2005.
[7] L.M. Fuentes and S.A. Velastin. "People Tracking in Surveillance Applications,” in Proc. of the IEEE Int'l Workshop Performance Evaluation of Tracking and Surveillance, pp. 20-27, Dec. 2001.
[8] Q. Zhou and J. Aggarwal, “Object tracking in an outdoor environment using fusion of features and cameras,” Image and Vision Computing, 24(11):1244–1255, 2006.
[9] S. Kong, M.K. Bhuyan, C. Sanderson, and B.C. Lovell, "Tracking of persons for video surveillance of unattended environments," in Proc. of the 19th Intern. Conf. on Pattern Recognition, pp. 1-4, December. 2008.
[10] P. KaewTraKulPong and R. Bowden, “An improved adaptive background mixture model for real-time tracking with shadow detection,” in Proc. 2nd European Workshop on Advanced Video Based Surveillance Systems, AVBS01, Sept 2001.
[11] P. W. Power and J. A. Schoonees, “Understanding background mixture models for foreground segmentation,” in Proc. Image and Vision Computing New Zealand, pp. 267-271, 2002.
[12] Z. Zivkovic, “Improved adaptive Gaussian mixture model for background subtraction,” in Proc. International Conference on Pattern Recognition, Vol.2, PP.28-31, Aug 2004.
[13] C. Anderson, Peter Burt, and G. van der Wal, “Change detection and tracking using pyramid transformation techniques”, in Proc. of SPIE - Intelligent Robots and Computer Vision, Vol. 579, pp. 72-78, 1985.
[14] A. AbuBaker, R. Qahwaji, S. Ipson, and M. Saleh, "One scan connected component labeling technique" in Proc. IEEE International Conference on Signal Processing and Communications, pp. 1283 – 1286, 2007.
[15] F. Chang, C. J. Chen, and C. J. Lu, “A linear-time component-labeling algorithm using contour tracing technique,” Comput. Vis. Image Understand, vol. 93, pp. 206–220, 2004.
[16] Grana, C., Borghesani, D., and Cucchiara, R., “Fast block based connected components labeling,” in Proc. IEEE International Conference on Image Process, Cairo, Egypt, Nov. 2009.
[17] Q. M. Li, and Z. Y. Zhang, “Improvement on adaptive mixture Gaussian background model,” Jisuanji Yingyong, Journal of Computer Applications. Vol. 27, no. 8, pp. 2014-2017. Aug. 2007.
[18] Yuh-Rau Wang, Wei-Hung Lin, and Shi-Jinn Horng, “A sliding window technique for efficient license plate localization based on discrete wavelete transform” , Expert Systems with Applications, vol. 38, issue 4 , pp.3142-3146, Apr. 2011.
[19] 林韋宏、謝家華、郭文德、賴岱佑等人著，數位多媒體系 統概論，文魁出版社， 2009年
[20] SULIMAN, C., CRUCERU, C., and MOLDOVEANU, F., “Kalman filter based tracking in an video surveillance system,” Advances in Electrical and Computer Engineering, vol.10, issue 2, pp.30-34, May 27-29, 2010.