本論文提出一遮蔽物件辨識系統，在系統上主要分成五個部份：1.背景濾 除[38]2.輪廓抽取 3.顯著點偵測法 4.區塊特徵抽取 5.區域性特徵辨識等步 驟。此系統是假設在能有效的擷取出前景的情況下工作，如此才能有效的觀察 遮蔽辨識系統的效率，採用了以 GMM 模組配合顏色以及梯度變化的背景濾除系 統找出較完整的前景。在論文中提出一有效的顯著點偵測法，此偵測法不需要 任何外來的輸入參數，可以工作在各種變化的輪廓上，且能克服些微雜訊所帶 來的影響以及利用較少的顯著點來表示完整的輪廓，其效果可參考第三章。在 特徵的選取上採用鄰近顯著點之間連線的角度與長度比當特徵，此特徵存在著 與大小、位移以及旋轉無關的特性。在第六章的實驗結果可看出此系統確實可 以解決位移、旋轉、遮蔽所帶來的問題，甚至能有效的標示出物件受遮蔽的部 份，在處理速度上達到平均每秒 20 張的速度，在速度上也接近實際時間 (real-time)的效果了。
一般的影像辨識系統主要分成 3 部份，影像擷取、特徵擷取以及特徵比對 辨識等三部份。但是本論文為了克服遮蔽帶來的影響，利用區域性資訊彼此不 相影響的優點取代全域性的特徵資訊當成解決遮蔽的方法，所以本系統比一般 辨識系統多了一個重要程序，那就是分割的動作，將一個全域特徵變成多個區 塊特徵，所以分割系統在整體系統效率上扮演著相當重要的角色，本論文提出 利用顯著點偵測法找出一個輪廓的顯著點進行分割的動作，所以此顯著點偵測 法可說是本系統最重要的一環，如果此系統能穩定且準確的找出各物件的顯著 點，那麼此辨識系統便會相當的完整有效率，但是此系統確實會被輪廓的變化 所影響(尤其是比例的變化)，導致顯著點偵測上有所偏差，進而破壞系統效率。
所以加強顯著點偵測法的穩定性以及效率是未來必要的，還有是否增加特徵來 解決比例變化所帶來的影響也是可以當作一個解決辦法，但是加特徵必定降低 系統處理速度，所以能在顯著點偵測法上有所改進是最有效果的。
參考文獻
[1] M. Sarfraz “Object Recognition using Fourier Descriptor: Some Experiments and Observations”, in Proc. IEEE Int. Conf. Computer Graphics, July 2006, 281-286 [2] Gorman, J.W.; Mitchell, O.R.; Kuhl, F.P., “Partial shape recognition using dynamic programming,” IEEE Trans. Pattern Analysis and Machine Intelligence. , Volume 10, March 1988, 257-266,
[3] U. Ramer, “An iterative procedure for the polygonal approximation of plane closed curves,”
Computer Graphics Image Processing, vol.1, 1972, 244-256
[4] M.P. Yu, K.C. Lo, “Object Recognition by Combining Viewpoint Invariant Fourier Descriptor and Convex Hull,” Video and Speech Processing, May 2001, 401-404
[5] N. Zheng, Y. Li, “Local Feature-Based Recognition of Partially Occluded Objects Using Neural Network,” IECON Proceedings (Industrial Electronics Conference), v 2, 1995, 1301-1306
[6] T. Du, K. B. Lim, G. S. Hong, W. M. Yu, H. Zheng, “2-D Occluded Object Recognition Using Wavelets,” Proceedings - The Fourth International Conference on Computer and Information Technology, 2004, 227-232
[7] C.Stauffer and W.Eric, L. Grimson,“Learning patterns of activity using real-time tracking”, IEEE Trans. on PAMI, 22(8) , Aug 2000, 747-757.
[8] H. Stark,J W. Woods, Probability and Random Processes with Applications to Signal Processing, 3th Edition , 2002
[9] T. Pavlidis, “Algorithms for Graphics and Image Processing,” Springer, Berlin, 1982 [10] A. Rosenfeld, E.Johnston, “Angle detection on digital curves,” IEEE Transactions on
Computers C-22 (1973) 875-878
[11] Wen-Yen Wu, “An adaptive method for detecting dominant points,” Pattern Recognition 36
[12] H. Freeman, “On the encoding of arbitrary geometric configurations,” IRE Transactions of Electronics Computing 10 (1961) 260-268
[13] M. Marji, P. Siy, “Polygonal representation of digital planar curves through dominant point detection – a nonparametric algorithm,” Pattern Recognition 37 (2004) 2113-2130
[14] C. The, R.Chin, “On the detection of dominant points on digital curves,” IEEE Transaction Pattern Analysis, Mach. Intell. 8 (1989) 859-872
[15] B.K. Ray, K.S. Ray, “Detection of significant points and polygonal approximation of digitized curves,” Pattern Recognition Lett. 12 (1992) 443-452
[16] W.Gorman, “partial shape recognition using dynamic programming,” IEEE transactions on pattern analysis and matching intelligence, 1988, 257-265
[17] S.Meikle, “Towards real-time object recognition using pairs of lines,” real time imaging, 2005, 31-43
[18]Hong-Chih Liu, “Partial Shape Classification using Contour Matching in Distance Transformation,” IEEE transactions on pattern analysis and matching intelligence, vol.12, no. 11, 1990
[19] S. H. Yoon, J. H. Kim, “An Optimum Solution for Scale-Invariant Object Recognition Based on the Multiresolution Approximation,” Pattern Recognition, vol. 31, No. 7, 1998, 889-908,
[20] Z. Ying, D. Castanon, “Feature Based Object Recognition using Statistical Occlusion Models with One-to-one Correspondence,” Proceedings of the IEEE International Conference on Computer Vision, v 1, 2001, 621-627
[21]Z. Ying, D. Castanon, “Partially Occluded Object Recognition Using Statistical Models,”
International Journal of Computer Vision, 2002, 57-78
[22]C. J. Chan, S. Y. Chen, “Recognizing Partially Occluded Objects Using Markov Model,”
International Journal of Pattern Recognition and Artificial Intelligence, vol.16, 2002, 161-191
[23]R. Dinesh, D. S. Guru, “Recognition of Partially Occluded Objects Using Perfect Hashing:
An Efficient and Robust Approach,” Proceedings of Conference on Computer and Robot Vision, May 2005, 528-535,
[24] M. P. Reinhold, M. Grzegorzek, “Appearance-based recognition of 3-D objects by cluttered background and occlusions,” Pattern Recognition Society, 2005, 739-753
[25] J. Kim, J. Choi, “Effective representation using ICA for face recognition robust to local distortion and partial occlusion,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol.27, 2005, 1977-1981
[26] Ray, Kumar S, G. Jayati, “Neuro-fuzzy reasoning for occluded object recognition,” Fuzzy Sets and Systems, Vol 94, February 1998, 1-28
[27] M. Zaki, R. Omran, “A soft computing approach for recognition of occluded shapes,” The Journal of System and Software, 2000, 73-83,
[28]P. Zhu, P.M. Chirlian, “On critical point detection of digital shapes,” IEEE Transactions of Pattern Analysis and Machine Intelligence 17 (8) (1995) 737-748
[29]N. Ansari, K.W. Huang, “non-parameter dominant point detection,” Pattern Recognition 24 (1991) 849-862
[30]B.K. Ray, R. Pandyan, “ACORD-an adaptive corner detector for planar curves,” Pattern Recognition 36 (2003) 703-708
[31]A. Garrido, N. Prez, M. Garca-Silvente, “Boundary simplification using amultiescale dominant-point detection algorithm,” Pattern Recognition 31 (1998) 791-804
[32]F. Mokhtarian, A.K. Mackworth, “A theory of multiscale-based shape representation for planar curves,” IEEE Transactions on Pattern Analysis and Machine Intelligence 14 (1992) 789-805
[33]P. Cornic, “Another look at dominant point detection of digital curves,” Pattern Recognition Letters 18 (1997) 13-25
[34]Y.H. Gu, T. Tjahijadi, “Coarse-to-fine planar object identification using invariant curve
[35]M. Marji, P. Siy, “A new algorithm for dominant point detection and polygonization of digital curves,” Pattern Recognition 36 (2003) 2239-2251
[36]C. Urdiales, A. Bandera, F. Sandoval, “Non-parameteric planar shape representation based on adaptive curvature functions,” Pattern Recognition 35 (2002) 43-53
[37] 林佩靜, 「以二維形狀特徵為基礎的人體姿勢辨識系統」, 國立交通大學,碩士論文,民 國95年
[38] 鄭士奇,「以高斯混合模型為基礎並使用陰影濾除之動態背景影像模型建立」,國立交通 大學,碩士論文,民國94年