使用高斯混合模型於移動物件偵測｜國立臺灣科技大學博碩士論文系統

簡易檢索 / 詳目顯示

回結果列表

研究生：	許瑞雄 Jui-Hsiang Hsu
論文名稱：	使用高斯混合模型於移動物件偵測 Moving Object Detection Using Gaussian Mixture Models
指導教授：	蘇順豐 Shun-Feng, Su
口試委員:	張志永 Jyh-Yeong Chang 姚立德 L. Yao 王偉彥 Wei-Yen Wang 蔡超人 Chau-Ren, Tsai
學位類別：	碩士 Master
系所名稱：	電資學院 - 電機工程系 Department of Electrical Engineering
論文出版年：	2007
畢業學年度：	95
語文別：	英文
論文頁數：	55
中文關鍵詞：	高斯混合模型、物件偵測、背影更新
外文關鍵詞：	Background update
相關次數：	點閱：177 下載：4
分享至:	分享至facebook 分享至twitter

查詢本校圖書館目錄查詢臺灣博碩士論文知識加值系統勘誤回報

上一筆

在本論文中，主要分成三個部分，首先，討論背景相減時所需的背景更新及相減後所需的門限方法，接著採用高斯混合模型(GMM)的方法建立動態的背景模型，並利用高斯分佈去估計模型的參數，接下來結合背景模型中每一點像素梯度的統計資訊，逐步地做前景與背景的判定，並由判定的結果與連續影像之間變化的程度，更新並重建背景的模型，讓模型具有記錄所有發生過狀況的能力。雖然此方法具有自動更新模型的參數及免除背景更新的優點，但因逐點計算耗時，且容易受樹葉等輕微移動，降低背景濾除的正確性，因此，我們提出一個以區塊取代像素的影像描述方法，能提高計算的速度、忽略輕微移動所造成的影響，最後，針對室外環境中可能發生的各種狀況，進行實驗驗證及討論。

This thesis consists of three parts. In the first part, the background subtraction method is introduced and the background update algorithm and the threshold method required in background subtraction are discussed. Next, a dynamic background image model based on the Gaussian Mixture Model method is introduced. In the approach, the model parameters are estimated by using Gaussian distributions. The gradient statistical information for each pixel is utilized to separate background and foreground. Although this method can effectively update the parameters and skip background, the computational time is large due to the pixel by pixel computation. Besides, small moving objects like leaf swaying under wind effects or noise will deteriorate the performance of system due to the pixel-wise sensitive. Therefore, we propose to use blocks instead of pixels as a unit for process. Such a simple modification can largely enhance the computation efficiency and neglects the influence caused by the migration of small moving objects. Besides, in order to improve the detection performance, we also propose to skip frames to enlarge the migration effects. The effects of those approaches are also analyzed. In our study, the methods are experimented in various outdoor conditions. Discussion and comparison on the results are given.

摘要.....................................................I
Abstract................................................II
致謝....................................................Ⅲ
Contents................................................Ⅳ
List of Tables..........................................Ⅵ
List of Figures.........................................Ⅶ
Chapter 1	Introduction
1.1	Motivation......................................1
1.2	Literature survey of related work...............2
1.3	Thesis overview.................................3
Chapter 2	Image processing and relate techniques
2.1	Background updating.............................5
2.1.1	  Histogram-based method........................5
2.1.2	  History-based background detection............6
2.2	Image segmentation techniques...................6
2.2.1	Motion detection................................7
2.2.2	Threshold method................................8
2.3	Median filter and Size filter..................10
2.3.1	Median filter..................................10
2.3.2	Size filter....................................10
2.4	Morphological..................................11
2.4.1	Mathematical Morphology........................12
2.4.2	Component labeling for clustering..............13
Chapter 3	Implementation
3.1	Background subtraction approach................16
3.2	Gaussian mixture model.........................19
3.3	Gaussian mixture method analysis and discussion.22
3.3.1	RGB and grayscale..............................22
3.3.2	Pixel, Weight and Standard Deviation...........25
Chapter 4	Experiments and Discussion
4.1	Block-Based Approach...........................30
4.2	Experiment Setup...............................31
4.3	Comparison of Various Algorithms...............32
4.4	Comparison of Different Block Sizes............36
4.5	Analysis of Skipping Frames....................43
4.6	Combination of Skipping Frames and Block-Based
         approach.......................................45
Chapter 5	Conclusions and Future Work....................51
Reference...............................................52
作者簡介................................................55

                                

[1] C. Stauffer and W. E. L. Grimson, “Adaptive background mixture models for real-time tracking,” 1999 IEEE Computer Society Conference on .Computer Vision and Pattern Recognition, Vol. 2 , June 1999.
[2] J. C. Tai and K. T. Song, “Background segmentation and its application to traffic monitoring using modified histogram,” 2004 IEEE International Conference on Networking, Sensing and Control, Vol. 1, March 2004.
[3] C. DeCoro, M. Burns and A. Misra, http://www.cs.princeton.edu/~cdecoro/traffic/
[4] R. C. Gonzalez and R. E. Woods. Digital image processing using Matlab, Pearson/Prentice Hall, 2004.
[5] C. Ridder, O. Munkelt and H. Kirchner, “Adaptive background estimation and foreground detection using Kalman-filtering,” Proceedings of International Conference on Recent Advances in Mechatronics, ICRAM'95, pp. 193-199, June 1995..
[6] N. Friedman and S. Russel, “Image segmentation in video sequences: A probabilistic approach,” Proc. of the Thirteenth Conference of Uncertainty in Artificial Intelligence, August 1997.
[7] Z. Zhu, B. Yang, G. Xu and D. Shi, “A real-time vision system for automatic traffic monitoring based on 2D spatio-temporal images,” Proceedings 3rd IEEE Workshop on Applications of Computer Vision, pp. 162-167, 1996.
[8] G. Bailo, M. Bariani, P. Ijas and M. Raggio, “Background estimation with Gaussian distribution for image segmentation, a fast approach,” Proceedings of the 2005 IEEE International Workshop on Measurement Systems for Homeland Security, Contraband Detection and Personal Safety Workshop, pp. 2-5, 2005.
[9] M. Isard and B. J. MacCormick, “A Bayesian multiple-blob tracker,” Proc. Int’l. Conf Computer Vision, pp. 34-41, July 2001,.
[10] S. Kamijo, “Traffic monitoring and accident detection at intersections,” IEEE Trans. Intelligent Transportation Systems, vol. 1. no. 2, pp. 108-118, June 2000.
[11] M. Karaman, L. Goldmann, D. Yu and T. Sikora, “Comparison of Static Background Segmentation Methods,” Department of Communication Systems, Technical University of Berlin, 2005.
[12] Y. Liu, S. Jiang, Q. Ye, W. Gao and Q. Huang, “Playfield Detection Using Adaptive GMM and Its Application,” In Proceedings of IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), vol. 2, pp. 421-424, Mar. 2005.
[13] H. Greenspan, J. Goldberger and A. Mayer, “Probabilistic Space-Time Video Modeling via Piecewise GMM,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 26, pp. 384-396, Mar. 2004.
[14] Z. Zivkovic, “Improved adaptive Gaussian mixture model for background subtraction,” in Proc. of IEEE International Conference on Pattern Recognition, vol. 2, pp. 28-31, 2004.
[15] D. S. Lee, J. Hull and B. Erol, “A Bayesian framework for Gaussian mixture background modeling,” in Proc. of IEEE International Conference on Image Processing, vol. 3, pp. 14-17, 2003
[16] M. Cristani, M. Bicego and V. Murino, “Integrated region- and pixel-based approach to background modelling,” in Proc. of IEEE Computer Society Conference on Motion and Video Computing, pp. 3-8, 2002.
[17] D. S. Lee, “Effective Gaussian mixture learning for video background subtraction,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 27. no. 5, pp. 827-832, 2005.
[18] L, Li and M. Lenug, ”Integrating intensity and texture differences for robust change detection,” IEEE Trans. On image processing, vol. 11. no2, pp.105-112, 2002.
[19] X. Song and G. Fan, ” Joint Key-Frame Extraction and Object Segmentation for Content-Based Video Analysis,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 16, pp. 904 – 914, July 2006.
[20] J. Zhang and D. Ma, “Nonlinear prediction for Gaussian mixture image models,” IEEE Transactions on Image Processing, vol. 13, pp. 836 – 847, June 2004.
[21] L. Liu and G. Fan,”Combined key-frame extraction and object-based video segmentation,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 15, pp. 869 – 884, July 2005.
[22] M. J. Carlotto, “A cluster-based approach for detecting man-made objects and changes in imagery,” IEEE Transactions on Geoscience and Remote Sensing, vol. 43, pp. 374 – 387, Feb 2005.
[23] K. Gilholm and D. Salmond, “Spatial distribution model for tracking extended objects,” IEE Proceedings Radar, Sonar and Navigation, vol. 152, pp. 364 – 371, October 2005.
[24] K.K. Paliwal and S. So, “Multiple frame block quantisation of line spectral frequencies using Gaussian mixture models,” 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing, Proceedings. vol. 1, May 2004.
[25] F. Y. Hu, Y. N. Zhang and Lan, ”Yao An effective detection algorithm for moving object with complex background,” 2005 International Conference on Machine Learning and Cybernetics, Vol. 8, Aug 2005.
[26] Y. L. Tian, M. Lu and Hampapur, “Robust and efficient foreground analysis for real-time video surveillance,” 2005 IEEE Computer Society Conference on a Computer Vision and Pattern Recognition, Vol.1, June 2005.

簡易檢索 / 詳目顯示

相關論文