Please use this identifier to cite or link to this item:
標題: 人體特徵值擷取及使用模糊類神經網路之姿態辨識
Human Body Feature Extraction And Posture Recognition Using Neural Fuzzy Network
作者: 李炯霖
Lee, Chiung-Ling
關鍵字: moving object segmentation;人體擷取;phuman body extraction;osture recognition;FNN;移動物體分割;姿態辨識;模糊類神經網路
出版社: 電機工程學系所
引用: [1] W. Hu, T. Tan, L. Wang, and S. Maybank, “A survey on visual surveillance of object motion and behaviors,” IEEE Transactions on System, Man, and Cybernetics, Part C: Applications and Reviews, vol.34, no.3, pp. 334-352, Aug. 2004. [2] I. Haritaoglu, D. Harwood and L. S. Davis, “ real-time surveillance of people and their activities,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 22, no. 8, pp. 809 - 830, Aug. 2000. [3] S. Y. Chen, S. Y. Ma, and L. G. Chen, “Efficient moving object segmentation algorithm using background registration technique,” IEEE Trans. Circuits and Systems for Video Technique, vol. 12, no. 7, pp. 577-586, July 2002. [4] C. Kim, J. Cho, and Y. Lee, “The relational properties among results of background subtraction”, Proc. of Int. Conf. on Advanced Communication Technology, vol. 3, pp. 1887-1890, 2008. [5] C. F. Juang, C. M. Chang, J. R. Wu, and D. M. Lee, “Computer vision-based human body segmentation and posture estimation,” IEEE Trans. Syst., Man, and Cyber., Part A: Systems and Humans, vol. 39, no. 1, pp. 119-133, Jan. 2009. [6] H.D. Cheng, X.H. Jiang, Y. Sun and J. Wang, “Color image segmentation: advances and prospects”, Pattern Recognition, vol. 34, pp. 2259-2281, 2001. [7] Q. Zhou and J.K. Aggarwal, “Tracking and classifying moving objects from video”, Proc. IEEE Int. Workshop Performance Evaluation of Tracking and Surveillance, pp. 52-59, Dec. 2001. [8] M. T. Razali and B. J. Adznan, “Detection and classification of moving object for smart vision sensor,” Proc. Information and Communication Technologies vol. 1, pp. 733-737, 2006. [9] T. C. Chen, Vision-based Real-Time 3D Human Posture Significant Points Estimation, Master Thesis, NCHU, Taiwan, R.O.C., 2009. [10] R. C. Gonzalez and R. E. Woods, Digital Image Processing 2/e, Prentice Hall, 2008 [11] A. F. Bobick and A. D. Wilson, “A state-based technique to the representation and recognition of gesture,” IEEE Transactions Pattern Analysis and Machine Intelligence, vol. 19, pp. 1325-1337, Dec. 1997. [12] T. Starmer, J. Weaver, and A. Pentland, “Real-time American sign language recognition using desk and wearable computer-based video,” IEEE Transactions Pattern Analysis Machine Intelligence, vol. 20, pp. 1317-1375, Dec. 1998. [13] N. Johnson and D. Hogg, “Learning the distribution of object trajectories for event recognition,” Image Vision Computer, vol. 14, no. 8, pp. 609-615, 1996. [14] H. Fujiyoshi and A. J. Lipton, “Real-time human motion analysis by image skeletonization,” Proceedings of IEEE Workshop on Applications of Computer Vision , pp. 15-21, Oct. 1998. [15] Y. Li, M. Songde, and L. Hanqing, “A multiscale morphological method for human posture recognition,” Proceedings of IEEE International Conference on Automatic Face and Gesture Recognition, pp. 56-61, Apr. 1998. [16] P. Spagnolo, M. Leo, A. Leone, G. Attolico, and A.Distante, “Posture estimation in visual surveillance of archaeological sites,” Proceedings of IEEE International Conference on Advanced Video and Signal Based Surveillance, pp. 277-283, July 2003. [17] C. F. Juang and C. M. Chang, “Human body posture classification by a neural fuzzy network and home care system application,” IEEE Trans. Syst., Man, and Cyber., Part A: Systems and Humans, vol. 37, no. 6, pp. 984-994, Nov. 2007. [18] V. Vapnik, The Nature of Statistical Learning Theory, Springer, New York, 1995. [19] E. Osuna, R. Freund and F. Girosi, “Training support vector machines: an application to face detection,” in: Proc. IEEE Int. Conf.Computer Vision and Pattern Recognition, pp. 130-136, 1997. [20] C. F. Juang, W. Y. Cheng, and T. C. Chen, “Batch support vector machine-trained fuzzy classifier with channel equalization application,” Proc. 5th IEEE Conf. Industrial Electronics and Applications, pp. 582-586, Taichung, Taiwan, June 2010. [21] M. Singh, A. Basu and M. Kr. Mandal, “Human activity recognition based on silhouette directionality,” IEEE Transactions on Circuits and Systems For Video Technology, vol.18, no. 9, pp. 1280-1292, Sep. 2008. [22] C. W. Hsu and C. J. Lin, “A comparison of methods for multiclass support vector machines,” IEEE Trans. Neural Networks, vol. 13, no. 2, pp. 415-525, Mar. 2002.
本論文提出了一種基於視覺的人體姿態辨識方法並採用模糊規則為基礎的分類器,目的在於辨識四種人體主要姿態,包含:站、彎、坐、躺。首先藉由兩台攝影機同時擷取兩組連續影像,運用以RGB三原色為基礎的移動物體分割演算法從背景中切割出人體,再配合一系列的影像處理方法,去除雜訊及陰影,最後得到兩組相對應的完整且乾淨之人體剪影。特徵值擷取部份,我們對剪影做水平與垂直投影後,經由離散傅立葉轉換(DFT)來取得投影向量的特徵值,並且配合人體的長寬比值,合併為一組特徵值。另外,將包住人體剪影的最小正方形網格均勻切割,求得剪影在每個分割網格的面積比例並以此作為特徵值,連同第一組特徵值代入分類器。本論文使用:自我建構類神經模糊推論網路(SONFIN)、使用支持向量機訓練之類神經模糊網路(SVM-trained NFN) 及使用高斯核之支持向量機(Gaussian-kernel-based SVM) 做姿態辨識。實驗結果顯示,三種辨識器皆有不錯的辨識效果,然而前兩種採用模糊規則為基礎的分類器所需的運算量較第三種少,更適合實際應用於即時辨識系統。

This thesis proposes a vision-based human posture recognition method using fuzzy rule-based classifiers. The objective is to recognize four kinds of main body postures, including standing, bending, sitting, and lying. This thesis first uses two cameras to capture two sets of image sequences at the same time. A RGB-based moving object segmentation algorithm is then proposed to distinguish the human body from background. After segmentation, a sequence of image processing approaches is used to reduce the influence of noise and shadow. Finally, two complete and corresponding silhouettes of the human body are obtained. For feature extraction, significant Discrete Fourier Transform (DFT) coefficients calculated from horizontal and vertical projections of each silhouette, together with length-width ratio of human body, are used as one set of features. Additionally, the minimum enclosing square of the body silhouette is uniformly partitioned and the area ratio of the silhouette in each partitioned grid is calculated used as another set of features. This thesis uses a Self-Constructing Neural Fuzzy Inference Network (SONFIN), a Support Vector Machine-trained fuzzy classifier (SVM-FC) and a Gaussian-kernel-based SVM to recognize the postures. Experimental results show that the three recognizers all have high recognition rates. However, the former two fuzzy rule-based classifiers show the advantage of smaller classifier size than the Gaussian-kernel-based SVM and are more suitable for practical implementation in a real-time recognition system.
其他識別: U0005-1908201020530800
Appears in Collections:電機工程學系所

Show full item record

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.