本研究提出基於顏色與目標物移動邊緣的改良型粒子濾波追蹤系統，在先行研究中，Nummiaro等人提出以顏色為基底的粒子濾波追蹤系統雖然可以有效的對目標物進行追蹤，但仍有一些限制，例如：目標物與背景顏色相似、目標物發生交錯遮蔽和目標物產生形變等狀況。因此為了改善上述限制，我們提出利用結合顏色和移動邊緣資訊的方法，使目標物與背景比對時系統能夠框選出較符合目標物大小的範圍框，這也確保後續目標物模型更新的正確性且在長時間的追蹤下能有更好的效能。實驗結果也證實，我們對100組目標物進行在室內室外數種環境下的影片作追蹤測試，本研究達到94.57 %的追蹤正確率，比Nummiaro等人的78.28 % 高得多，而在目標物發生交錯遮蔽的情況下，本研究也比Nummiaro等人的高出24.22%的追蹤正確率( 91.81 % - 67.59 % )，這說明我們的方法在一般和遮蔽的情況下，都具有更好的追蹤強鍵性。

In this thesis, an improved particle filtering tracking method combining color information and moving edge of target object was proposed. The previous work of Nummiaro et al. [9] uses color model to track object can achieve tracking efficiently. However, a few cases may not be handled by such a method, e.g., target and background with similar color, occlusion, and deformation of object’s shape, and it results in poor results. To solve this problem, this study used the information of moving edge of target object to ensure the target can be well enclosed by bounding box. This can also ensure correctness of the subsequent updating of target model, and prove better performance in long term tracking. Experiment results show the proposed method can achieve the correct rate 94.57%, which is much higher than the correct rate 78.28 % achieved by Nummiaro et al. when tracking 100 target objects in indoor and outdoor video sequences. For the case of occlusion, the proposed method can obtain more than 24.22% (91.81 % - 67.59 %) correct rate than the method proposed by Nummiaro et al. This shows our method can achieve more robust tracking correctness in both general and occlusion cases.

[1] I. Haritaoglu, D. Harwood and L. S. Davis, “W4: A Real Time System for Detecting and Tracking People,” cvpr, pp.962, 1998 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'98), 1998
[2] S. Huwer and H. Niemann, “Adaptive Change Detection for Real-Time Surveillance Applications,” Proceedings of IEEE third International Workshop on Visual Surveillance, pp.37-46, July 2000.
[3] Y. Ran and Q. Zheng, “multi moving people detection from binocular sequences,” Proceedings of IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 3, pp. 37-40, Apr. 2003.
[4] A. Talukder and L. Matthies, “Real-Time Detection of Moving Objects from Moving Vehicles Using Dense Stereo and Optical Flow,” Proceedings of IEEE/ESJ International Conference on Intelligent Robots and systems, vol. 4, pp. 3718-3725, Oct. 2004.
[5] F. E. Alsaqre, and Yuan Baozong, “Moving object segmentation from video sequences: An edge approach,” 4th EURASIP Conference on Video/Image Processing and Multimedia Comm., vol. 1, pp. 193-199, July 2003.
[6] T. Horprasert, D. Harwood, and L. S. Davis, “A Statistical Approach for Real-time Robust Background Subtraction and Shadow Detection,” Proceedings of IEEE ICCV’99 FRAME-RATE Workshop, 1999.
[7] J. Lim, D. Kriegman, “Tracking Humans using Prior and Learned Representations of Shape and Appearance,” Proceedings of IEEE International Conference on Automatic Face and Gesture Recognition(FGR’04), 2004.
[8] S. A. El-Azim, I. Ismail, and H. A. E1-Latiff, “An Efficient Object Tracking Technique using Block-matching Algorithm,” Proceedings of the Nineteenth National, Radio Science Conference, pp. 427-433, 2002.
[9] K. Nummiaro, E. Koller-Meier, L. Van Gool, “An adaptive color-based particle filter,” Image and Vision Computing 21, pp. 99-100, 2003.
[10] M. Heikkila, M. Pietikainen, "A Texture-Based Method for Modeling the Background and Detecting Moving Objects," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 28, no. 4, pp. 657-662, April 2006, doi:10.1109/TPAMI.2006.68
[11] D. Comaniciu, V. Ramesh, P. Meer, “Real-time Tracking of Non-rigid Objects Using Mean Shift,” Proceedings of IEEE Computer Vision and Pattern Recognition, pp. 142-149, 2000.
[12] Chris Stauffer, W.E.L Grimso “Adaptive background mixture models for real-time tracking” Proceedings 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Cat No PR00149 (1999)
[13] G.Welch and G. Bishop, “An Introduction to the Kalman Filter,”Apr.2004.
[14] Kenji Okuma, Ali Taleghani, Nando de Freitas, Jim Little, and David Lowe. ”A Boosted Particle Filter: Multitarget Detection and Tracking,” 8th European Conference on Computer Vision, ECCV 2004.
[15] P. Pérez, C. Hue, J. Vermaak, and M. Gangent, “Color-Based Probabilistic Tracking,” European Conference on Computer Vision, 2002.
[16] E. B. Meier and F. Ade, “Using the Condensation Algorithm to Implement Tracking for Mobile Robots,”1999 Third European Workshop on 6-8 Advanced Mobile Robots, pp.73-80, Sept. 1999.
[17] 宋炫慶，范欽雄，「基於粒子濾波技術的多個移動物體之即時是絕偵測與 追蹤」，國立台灣科技大學，民國94年。
[18] 陳郁龍，江政欽，「改良連續蒙特卡羅法於視訊物件追蹤之研究」，國立東 華大學，民國95年。